Nabble - BioPerl

tag:old.nabble.com,2006:forum-13596 Nabble - BioPerl 2009-11-26T03:27:50Z Officially organized in 1995 and existing informally for several years prior, The Bioperl Project is an international association of developers of open source Perl tools for bioinformatics, genomics and life science research. BioPerl home is <a href="http://bio.perl.org/" target="_top" rel="nofollow">here</a>. tag:old.nabble.com,2006:post-26528082 Re: Function that determines serious mutations 2009-11-26T03:27:50Z 2009-11-26T03:27:50Z RobertBradbury On Thu, Nov 26, 2009 at 5:54 AM, Alden Huang <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26528082&i=0" target="_top" rel="nofollow">alden.huang@...</a>> wrote: <br>> <br>> Sorting Intolerant from Tolerant <br>> <a href="http://sift.jcvi.org/" target="_top" rel="nofollow">http://sift.jcvi.org/</a><br>> <br>> <br>Ah yes, thank you very much.  This looks very much like a tool that can be <br>adapted for various uses. <br><br>Robert <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26528082&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26527689 Re: Function that determines serious mutations 2009-11-26T02:54:30Z 2009-11-26T02:54:30Z Alden Huang Hey rob, <br><br>Sorting Intolerant from Tolerant <br><a href="http://sift.jcvi.org/" target="_top" rel="nofollow">http://sift.jcvi.org/</a><br><br>~alden <br><br>...a bit late, i kno; I just read you post now while cleaning the inbox <br><br>On Fri, Nov 6, 2009 at 9:35 AM, Robert Bradbury <br><<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26527689&i=0" target="_top" rel="nofollow">robert.bradbury@...</a>> wrote: <div class='shrinkable-quote'><br>> Is there a function in the library (or has someone written one) that can <br>> take a genbank entry and determine which mutations are harmful? <br>> <br>> It would be used to produce a table summary of: <br>> GENE # SNP # BadSNP <br>> <br>> One kind of gets this from NCBI if you lookup in the "GENE" db a gene name <br>> and then go to the "GeneView" om dbSNP page it has the information I want <br>> but largely in a graphical format while I simply want numbers I can dump <br>> into a spreadsheet. <br>> <br>> I don't think it would be hard, fetch the gene, run through the features for <br>> the SNP database, figure out whether they are good or bad SNPs, accumulate <br>> the statistics and dump it. I think the functions available are flexible <br>> enough to do it but I can't believe nobody has already done it. It could be <br>> a bit more complex in that one could do an analysis to see if the mutations <br>> are in a conserved domain or mutations that code for Cysteine or Methionine <br>> (or othe potentially "critical" amino acids) but since "critical" is in the <br>> eye of the beholder there would have to be some kind of callback to a <br>> scoring function. <br>> <br>> Thanks, <br>> Robert <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26527689&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>> </div><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26527689&i=2" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26518941 Re: How to parse BLAST output - all hits of each queryinnew file 2009-11-25T11:21:27Z 2009-11-25T11:21:27Z Mark A. Jensen whoops: change the following line: <br>my $blio = Bio::SearchIO->new( -file => $qid.".bls", -format=>'blast' ); <br><br>to <br><br>my $blio = Bio::SearchIO->new( -file => ">$qid\.bls", -format=>'blast' ); <br><br>(I always forget that...) <br>MAJ <br><br>----- Original Message ----- <br>From: "Mark A. Jensen" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518941&i=0" target="_top" rel="nofollow">maj@...</a>> <br>To: "Tim" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518941&i=1" target="_top" rel="nofollow">timbourine81@...</a>>; <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518941&i=2" target="_top" rel="nofollow">bioperl-l@...</a>> <br>Sent: Wednesday, November 25, 2009 1:20 PM <br>Subject: Re: [Bioperl-l] How to parse BLAST output - all hits of each queryinnew <br>file <br><br><div class='shrinkable-quote'><br>> hey Tim-- <br>> <br>> Sound like you need to go about collecting your queries inside out: <br>> <br>> my %hits_by_query; <br>> for ($result->hits) { <br>>  push @{$hits_by_query{$hit->name}} $hit; <br>> } <br>> <br>> I believe now each hash element, keyed by the query name, will contain <br>> an arrayref to the set of hits assoc with that query. <br>>>From here, I believe <br>> <br>> use Bio::Search::Result::BlastResult; <br>> use Bio::SearchIO; <br>> <br>> foreach my $qid ( keys %hits_by_query ) { <br>>  my $result = Bio::Search::Result::BlastResult->new(); <br>>  $result->add_hit($_) for ( @{$hits_by_query{$qid}} ); <br>>  my $blio = Bio::SearchIO->new( -file => $qid.".bls", -format=>'blast' ); <br>>  $blio->write_result($result); <br>> } <br>> <br>> will do what you want. <br>> <br>> hope this helps - <br>> Mark <br>> <br>> ----- Original Message ----- <br>> From: "Tim" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518941&i=3" target="_top" rel="nofollow">timbourine81@...</a>> <br>> To: <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518941&i=4" target="_top" rel="nofollow">bioperl-l@...</a>> <br>> Sent: Wednesday, November 25, 2009 12:40 PM <br>> Subject: [Bioperl-l] How to parse BLAST output - all hits of each query innew <br>> file <br>> <br>> <br>>> Dear bioperl users, <br>>> <br>>> I am a real newbie and have - maybe a very trivial - question. <br>>> <br>>> I searched the mailing list archive and many howtos but I have not found <br>>> a concrete answer to my problem. So hopefully you can help me :) <br>>> <br>>> Background: I use the latest Bioperl version (installed it two weeks <br>>> before). <br>>> When I use Bio::Tools::Run::StandAloneBlast to BLAST one fasta file <br>>> including different sequences, I get a BLAST output with many queries <br>>> each having several hits / sbjcts. <br>>> <br>>> My problem is how to parse *all* hits of *one* query into a single new <br>>> file. And this for all the queries I have in my BLAST output file. <br>>> <br>>> Or is it better the other way round; first to make fasta files with only <br>>> single sequences inside and BLAST each file? But how can I automize that <br>>> using Bioperl? <br>>> <br>>> I tried Bio::SearchIO but can only parse all queries and their <br>>> respective hits in only one file... <br>>> I think iteration is also necessary here, but I do not really know how <br>>> to include that into Bio::SearchIO. <br>>> Or do I have to use Module:Bio::Index::Blast? <br>>> <br>>> I can index a file (see below), but I have no idea what comes next... <br>>> <br>>> ###How I index a file... <br>>> <br>>> #!/usr/bin/perl -w <br>>> <br>>> $ENV{BIOPERL_INDEX_TYPE} = "SDBM_File"; <br>>> <br>>> use Bio::Index::Fasta; <br>>> <br>>> <br>>> $file_name = "8_to_BLAST_two_seq_index.fasta"; <br>>> $id = "48882"; <br>>> $inx = Bio::Index::Fasta->new (-filename => $file_name . ".idx", <br>>> -write_flag => 1); <br>>> $inx->make_index($file_name); <br>>> <br>>> <br>>> Hopefully, you can give me at least hints what to look for. <br>>> <br>>> A big THANKS in advance! <br>>> <br>>> Cheers, <br>>> <br>>> Tim <br>>> _______________________________________________ <br>>> Bioperl-l mailing list <br>>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518941&i=5" target="_top" rel="nofollow">Bioperl-l@...</a> <br>>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>>> <br>>> <br>> <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518941&i=6" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>> <br>> </div><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518941&i=7" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26518665 Re: How to parse BLAST output - all hits of each query in new file 2009-11-25T11:07:26Z 2009-11-25T11:07:26Z Smithies, Russell Hi Tim, <br>Here's some code for a job I'm working on at the moment that contains all the bits you'll probably need. <br>It's extracting 2 species-specific databases from nr (based on tax ids), doing a blast, then parsing the results and creating a substitution matrix. I was initially using Bio::DB::Eutilities to query and retrieve sequences but I kept getting errors and time-outs from NCBI when pulling back large numbers of sequences. <br>It should give you a rough idea of how to run Bio::Tools::Run::StandAloneBlast, Bio::DB::Fasta and Bio::SearchIO. <br><br>Email me direct if you want further explaination as it's not well commented ;-) <br><br>Russell Smithies <br><br>Bioinformatics Applications Developer <br>T +64 3 489 9085 <br>E  <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518665&i=0" target="_top" rel="nofollow">russell.smithies@...</a> <br><br>Invermay  Research Centre <br>Puddle Alley, <br>Mosgiel, <br>New Zealand <br>T  +64 3 489 3809   <br>F  +64 3 489 9174   <br>www.agresearch.co.nz <br><br>======================================= <br><br>#!/usr/local/bin/perl <br><br>use strict; <br>use Bio::Tools::Run::StandAloneBlast; <br>use Bio::SeqIO; <br>use Bio::SearchIO; <br>use Bio::DB::Fasta; <br><br>use Storable; <br><br># Parameters: <query> <subject> <number or percentage of searches> <br># Percentage can be specified as either 20p, 20P or 20% <br># So for 20% of rice sequences blasted against oil palm: <br>#    4530 51953 20p   (4530=rice,51953=oil_palm, 20p=20%) <br># Or for 20 searches: <br>#      4530 51953 20 <br># <br>my ( $q, $s, $c ) = @ARGV; <br><br>my $nr = "/data/databases/flatfile/illuminati_blastdata/nr"; <br>my $tax_file = "/data/anonftp/pub/mirror/taxonomy/gi_taxid_prot.dmp.gz"; <br>my $tmp = "/tmp/tax"; <br><br><br>my %stats      = (); <br>my $total_subs = 0; <br><br>my $min_hsp_len      = 0; <br>my $min_hsp_identity = 0; <br>my $num_searches     = $c || 10; <br>my $blast_e          = '1e-6'; <br>my $count            = 0; <br><br># check if all the fasta and blast files exist <br># if not, extract new fasta and re-formatdb the database <br>foreach my $t ( $q, $s ) { <br>  foreach ( map { "$tmp/$t.$_" } qw(faa list phr pin psq) ) { <br>    unless ( -e $_ ) { <br>      print "Creating database for $t\n"; <br>      &create_database($t); <br>      last; <br>    } <br>  } <br>} <br><br>my @params = ( <br>               -database => "$tmp/$q", <br>               -program  => 'blastp', <br>               -e        => $blast_e, <br>               -outfile  => "$tmp/blast.out", <br>               -v        => '1', <br>               -b        => '1' <br>); <br>my $factory = Bio::Tools::Run::StandAloneBlast->new(@params) or die $!; <br><br># load the query sequences into a db <br># makes it easier to randomly access them <br>my $db = Bio::DB::Fasta->new( "$tmp", -glob => "$s.faa", -reindex => 1 ); <br><br>my @ids      = $db->ids; <br>my $id_count = $#ids; <br>exit "No sequences\n" unless $id_count; <br><br># if a percentage is requested, calculate <br># the required number of searches <br>if ( $num_searches =~ m/(\d+)[pP%]/ ) { <br>  $num_searches = int( ( $1 / 100 ) * $id_count ); <br>  warn <br>"Searching random $1 percent ($num_searches) of $id_count sequences from taxid $q\n"; <br>} <br><br>my $summary_file = "$tmp/".$$."_summary.txt"; <br>open( OUT, ">", $summary_file ) or die $!; <br>print OUT <br>"#Summary of $num_searches random blast searches from taxid $q against taxid $s.\n"; <br>print OUT "#Parameters used were:\n"; <br>print OUT "#blast_e: $blast_e\n"; <br>print OUT "#min_hsp_len: $min_hsp_len\n"; <br>print OUT "#min_hsp_identity: $min_hsp_identity\n"; <br>print OUT "\n"; <br><br>while ( my $seq = $db->get_Seq_by_id( $ids[ rand($#ids) ] ) ) { <br>  next unless $seq; <br><br>  warn "Processing ", $seq->id, "\n"; <br>  eval { <br>    my $blast_report = $factory->blastall($seq); <br>    sleep 5; <br>  }; <br><br>  my $blast_in = new Bio::SearchIO( -format => "blast", -file => "$tmp/blast.out" ); <br><br>  while ( my $result = $blast_in->next_result ) { <br>    if ( $result->num_hits <= 0 ) { <br>      warn "No hits for ", $result->query_accession, "\n"; <br>      print OUT "No hits for ", $result->query_accession, "\n"; <br>      next; <br>    } <br>    $count++; <br>    while ( my $hit = $result->next_hit ) { <br>      while ( my $hsp = $hit->next_hsp ) { <br>        warn sprintf( "%s had %s hsp%s\n", <br>                      $result->query_accession, $hit->num_hsps, <br>                      $hit->num_hsps > 1 ? "s" : "" ); <br>        print OUT sprintf( "%s had %s hsp%s\n", <br>                      $result->query_accession, $hit->num_hsps, <br>                      $hit->num_hsps > 1 ? "s" : "" ); <br><br>        # <a href="http://www.bioperl.org/wiki/HOWTO:SearchIO#Table_of_Methods" target="_top" rel="nofollow">http://www.bioperl.org/wiki/HOWTO:SearchIO#Table_of_Methods</a><br>        if ( $hsp->length('total') > $min_hsp_len ) { <br>          if ( $hsp->percent_identity >= $min_hsp_identity ) { <br>            my @query_string = split '', $hsp->query_string; <br>            my @homol_string = split '', $hsp->homology_string; <br>            my @hit_string   = split '', $hsp->hit_string; <br>            for ( my $i = 0; $i < $#query_string; $i++ ) { <br>              next unless $homol_string[$i] =~ /\+/; <br>              $stats{ $query_string[$i] }{ $hit_string[$i] }++; <br>              $total_subs++; <br>            } <br>          } <br>        } <br>      } <br>    } <br>  } <br>  unlink '$tmp/blast.out' if -e '$tmp/blast.out'; <br>  last if $count >= $num_searches; <br>} <br><br><br><br># create summary frequency list <br>my %summary = (); <br>for my $query ( keys %stats ) { <br>  for my $hit ( keys %{ $stats{$query} } ) { <br>    $summary{"$query->$hit"} = <br>      sprintf( "%6f", $stats{$query}{$hit} / $total_subs ); <br>  } <br>} <br><br>print OUT "\n"; <br><br># sort by decending frequencies and print to summary file <br>foreach my $k ( sort { $summary{$b} <=> $summary{$a} } keys %summary ) { <br>  print OUT "$k\t", $summary{$k}, "\n" unless $k =~ /TOTAL/; <br>} <br><br>print OUT "\n\n"; <br><br># print substitution matrix <br>my $i     = 0; <br>my @prots = qw(A R N D C Q E G H I L K M F P S T W Y V); <br>my $sep   = "\t"; <br><br>print OUT sprintf( "%7s %s", $_, $sep ) foreach ( "       ", @prots ); <br>print OUT "\n"; <br><br>foreach my $x (@prots) { <br>  print OUT sprintf( "%7s|%s", $prots[ $i++ ], $sep ); <br>  foreach my $y (@prots) { <br>    my $val = <br>      defined( $stats{$x}{$y} ) <br>      ? sprintf( "%0.6f", $stats{$x}{$y} / $total_subs ) <br>      : "--------"; <br>    print OUT sprintf( "%s%s", $val, $sep ); <br>  } <br>  print OUT "\n"; <br>} <br>close OUT; <br><br><br>open(IN, $summary_file) or die $!; <br>print $_ while(<IN>); <br>close IN; <br><br><br># extract sequences from nr database based on taxid. <br>sub create_database { <br>  my $txid      = shift; <br>  my %hash      = (); <br>  my $gi_stored = "/tmp/gi.dat"; <br><br>  if ( -e $gi_stored ) { <br>    %hash = %{ retrieve($gi_stored) }; <br>  } <br>  else { <br>    open( TXID, "zcat $tax_file | " ) or die $!; <br>    while (<TXID>) { <br>      chomp; <br>      my ( $gi, $tx ) = split( "\t", $_ ); <br>      push( @{ $hash{$tx} }, $gi ); <br>    } <br>    close TXID; <br><br>    store( \%hash, $gi_stored ); <br>  } <br><br>  my $txlist = "$tmp/$txid.list"; <br>  my $txseq  = "$tmp/$txid.faa"; <br>        <br>        die "No sequences found for taxid $txid\n" unless defined( @{ $hash{$txid} }); <br>        my $num_seqs =  scalar( @{ $hash{$txid} }); <br>        warn "Found $num_seqs sequences for taxid $txid in $tax_file\n"; <br><br>  open OUT, ">", $txlist or die $!; <br>  print OUT "$_\n" foreach ( @{ $hash{$txid} } ); <br>  close OUT; <br><br>  my $cmd = "fastacmd -d $nr -i $txlist -t T -o $txseq 2>/dev/null"; <br>  system $cmd; <br><br>  my $count = `grep -c '>' $txseq`; <br>  $count =~ s/\n//; <br>        warn "Could only extract $count sequences from $nr\n"; <br><br>  $cmd = "formatdb -p T -i $tmp/$txid.faa -n $tmp/$txid -l $tmp/formatdb.log"; <br>  system $cmd; <br><br>  $cmd = "fastacmd -d $tmp/$txid -I"; <br>  system $cmd; <br><br>  warn "Check the formatdb.log for any errors\n"; <br>} <br><br><br>======================================= <br><br><br><br><br><br><div class='shrinkable-quote'><br>> -----Original Message----- <br>> From: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518665&i=1" target="_top" rel="nofollow">bioperl-l-bounces@...</a> [mailto:bioperl-l- <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518665&i=2" target="_top" rel="nofollow">bounces@...</a>] On Behalf Of Tim <br>> Sent: Thursday, 26 November 2009 6:41 a.m. <br>> To: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518665&i=3" target="_top" rel="nofollow">bioperl-l@...</a> <br>> Subject: [Bioperl-l] How to parse BLAST output - all hits of each query in <br>> new file <br>> <br>> Dear bioperl users, <br>> <br>> I am a real newbie and have - maybe a very trivial - question. <br>> <br>> I searched the mailing list archive and many howtos but I have not found <br>> a concrete answer to my problem. So hopefully you can help me :) <br>> <br>> Background: I use the latest Bioperl version (installed it two weeks <br>> before). <br>> When I use Bio::Tools::Run::StandAloneBlast to BLAST one fasta file <br>> including different sequences, I get a BLAST output with many queries <br>> each having several hits / sbjcts. <br>> <br>> My problem is how to parse *all* hits of *one* query into a single new <br>> file. And this for all the queries I have in my BLAST output file. <br>> <br>> Or is it better the other way round; first to make fasta files with only <br>> single sequences inside and BLAST each file? But how can I automize that <br>> using Bioperl? <br>> <br>> I tried Bio::SearchIO but can only parse all queries and their <br>> respective hits in only one file... <br>> I think iteration is also necessary here, but I do not really know how <br>> to include that into Bio::SearchIO. <br>> Or do I have to use Module:Bio::Index::Blast? <br>> <br>> I can index a file (see below), but I have no idea what comes next... <br>> <br>> ###How I index a file... <br>> <br>> #!/usr/bin/perl -w <br>> <br>> $ENV{BIOPERL_INDEX_TYPE} = "SDBM_File"; <br>> <br>> use Bio::Index::Fasta; <br>> <br>> <br>> $file_name = "8_to_BLAST_two_seq_index.fasta"; <br>> $id = "48882"; <br>> $inx = Bio::Index::Fasta->new (-filename => $file_name . ".idx", <br>> -write_flag => 1); <br>> $inx->make_index($file_name); <br>> <br>> <br>> Hopefully, you can give me at least hints what to look for. <br>> <br>> A big THANKS in advance! <br>> <br>> Cheers, <br>> <br>> Tim <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518665&i=4" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a></div>======================================================================= <br>Attention: The information contained in this message and/or attachments <br>from AgResearch Limited is intended only for the persons or entities <br>to which it is addressed and may contain confidential and/or privileged <br>material. Any review, retransmission, dissemination or other use of, or <br>taking of any action in reliance upon, this information by persons or <br>entities other than the intended recipients is prohibited by AgResearch <br>Limited. If you have received this message in error, please notify the <br>sender immediately. <br>======================================================================= <br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518665&i=5" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26518581 Re: How to parse BLAST output - all hits of each query innew file 2009-11-25T10:20:03Z 2009-11-25T10:20:03Z Mark A. Jensen hey Tim-- <br><br>Sound like you need to go about collecting your queries inside out: <br><br>my %hits_by_query; <br>for ($result->hits) { <br>  push @{$hits_by_query{$hit->name}} $hit; <br>} <br><br>I believe now each hash element, keyed by the query name, will contain <br>an arrayref to the set of hits assoc with that query. <br>>From here, I believe <br><br>use Bio::Search::Result::BlastResult; <br>use Bio::SearchIO; <br><br>foreach my $qid ( keys %hits_by_query ) { <br>  my $result = Bio::Search::Result::BlastResult->new(); <br>  $result->add_hit($_) for ( @{$hits_by_query{$qid}} ); <br>  my $blio = Bio::SearchIO->new( -file => $qid.".bls", -format=>'blast' ); <br>  $blio->write_result($result); <br>} <br><br>will do what you want. <br><br>hope this helps - <br>Mark <br><br>----- Original Message ----- <br>From: "Tim" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518581&i=0" target="_top" rel="nofollow">timbourine81@...</a>> <br>To: <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518581&i=1" target="_top" rel="nofollow">bioperl-l@...</a>> <br>Sent: Wednesday, November 25, 2009 12:40 PM <br>Subject: [Bioperl-l] How to parse BLAST output - all hits of each query innew <br>file <br><br><div class='shrinkable-quote'><br>> Dear bioperl users, <br>> <br>> I am a real newbie and have - maybe a very trivial - question. <br>> <br>> I searched the mailing list archive and many howtos but I have not found <br>> a concrete answer to my problem. So hopefully you can help me :) <br>> <br>> Background: I use the latest Bioperl version (installed it two weeks <br>> before). <br>> When I use Bio::Tools::Run::StandAloneBlast to BLAST one fasta file <br>> including different sequences, I get a BLAST output with many queries <br>> each having several hits / sbjcts. <br>> <br>> My problem is how to parse *all* hits of *one* query into a single new <br>> file. And this for all the queries I have in my BLAST output file. <br>> <br>> Or is it better the other way round; first to make fasta files with only <br>> single sequences inside and BLAST each file? But how can I automize that <br>> using Bioperl? <br>> <br>> I tried Bio::SearchIO but can only parse all queries and their <br>> respective hits in only one file... <br>> I think iteration is also necessary here, but I do not really know how <br>> to include that into Bio::SearchIO. <br>> Or do I have to use Module:Bio::Index::Blast? <br>> <br>> I can index a file (see below), but I have no idea what comes next... <br>> <br>> ###How I index a file... <br>> <br>> #!/usr/bin/perl -w <br>> <br>> $ENV{BIOPERL_INDEX_TYPE} = "SDBM_File"; <br>> <br>> use Bio::Index::Fasta; <br>> <br>> <br>> $file_name = "8_to_BLAST_two_seq_index.fasta"; <br>> $id = "48882"; <br>> $inx = Bio::Index::Fasta->new (-filename => $file_name . ".idx", <br>> -write_flag => 1); <br>> $inx->make_index($file_name); <br>> <br>> <br>> Hopefully, you can give me at least hints what to look for. <br>> <br>> A big THANKS in advance! <br>> <br>> Cheers, <br>> <br>> Tim <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518581&i=2" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>> <br>> </div><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26518581&i=3" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26517465 How to parse different (fasta) files 2009-11-25T09:53:34Z 2009-11-25T09:53:34Z Tim-441 Hey everybody, <br><br>another question from me...if you do not mind :) <br><br>My situation is like this: I have parsed a standalone BLAST output using <br>SearchIO with only the hit names. Now I have a second fasta file with <br>the same sequences like in the BLAST database but including an alignment <br>(meaning "." and "-"). (There is no chance to make a BLAST database with <br>fasta files including the alignment, unfortunately...). <br>My intention is now to take the name of the hit sequences (BLAST output) <br>and to get the corresponding aligned sequences (fasta file incl. <br>alignment) and putting it in a new file. <br><br>Is anybody out there who has tried that before? <br><br>Again, I am a absolute greenhorn in using (Bio)perl. Maybe it is very <br>simple :D <br><br>Looking forward to get an answer of you. <br><br>All the best, <br><br>Tim <br>-- <br>Tim Köhler <br>MPI for Terrestrial Microbiology <br>Karl-von-Frisch-Straße <br>D-35043 Marburg / Germany <br><br>Email: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26517465&i=0" target="_top" rel="nofollow">koehlerd@...</a> <br>Phone: +49 6421 178-740 <br>Fax:   +49 6421 178-999 <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26517465&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26517377 How to parse BLAST output - all hits of each query in new file 2009-11-25T09:40:52Z 2009-11-25T09:40:52Z Tim-441 Dear bioperl users, <br><br>I am a real newbie and have - maybe a very trivial - question. <br><br>I searched the mailing list archive and many howtos but I have not found <br>a concrete answer to my problem. So hopefully you can help me :) <br><br>Background: I use the latest Bioperl version (installed it two weeks <br>before). <br>When I use Bio::Tools::Run::StandAloneBlast to BLAST one fasta file <br>including different sequences, I get a BLAST output with many queries <br>each having several hits / sbjcts. <br><br>My problem is how to parse *all* hits of *one* query into a single new <br>file. And this for all the queries I have in my BLAST output file. <br><br>Or is it better the other way round; first to make fasta files with only <br>single sequences inside and BLAST each file? But how can I automize that <br>using Bioperl? <br><br>I tried Bio::SearchIO but can only parse all queries and their <br>respective hits in only one file... <br>I think iteration is also necessary here, but I do not really know how <br>to include that into Bio::SearchIO. <br>Or do I have to use Module:Bio::Index::Blast? <br><br>I can index a file (see below), but I have no idea what comes next... <br><br>###How I index a file... <br><br>#!/usr/bin/perl -w <br><br>$ENV{BIOPERL_INDEX_TYPE} = "SDBM_File"; <br><br>use Bio::Index::Fasta; <br><br><br>$file_name = "8_to_BLAST_two_seq_index.fasta"; <br>$id = "48882"; <br>$inx = Bio::Index::Fasta->new (-filename => $file_name . ".idx", <br>-write_flag => 1); <br>$inx->make_index($file_name); <br><br><br>Hopefully, you can give me at least hints what to look for. <br><br>A big THANKS in advance! <br><br>Cheers, <br><br>Tim <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26517377&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26516981 question for all regarding a sam-based Bio::Assembly::IO 2009-11-25T09:25:30Z 2009-11-25T09:25:30Z Mark A. Jensen Short-readers, <br><br>I'm working on an Assembly::IO class for sam alignments. <br>I'm currently making a decision about handling multiple reference sequences: <br>would you prefer that next_assembly() return an assembly that covers all reference <br>sequences, or that next_assembly iterates over each reference sequence? <br>(Or both?) <br><br>thanks for your input- <br>MAJ <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26516981&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26507326 Re: Bio::DB::Fasta 2009-11-24T20:12:13Z 2009-11-24T20:12:13Z Mark A. Jensen I seem to be able to do <br>$db = Bio::DB::Fasta->new("$tmp/test.faa"); <br>without a problem- something in the mixing of named and unnamed parameters? <br>----- Original Message ----- <br>From: "Smithies, Russell" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26507326&i=0" target="_top" rel="nofollow">Russell.Smithies@...</a>> <br>To: "'Mark A. Jensen'" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26507326&i=1" target="_top" rel="nofollow">maj@...</a>>; "'bioperl-l'" <br><<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26507326&i=2" target="_top" rel="nofollow">bioperl-l@...</a>> <br>Sent: Tuesday, November 24, 2009 5:21 PM <br>Subject: RE: [Bioperl-l] Bio::DB::Fasta <br><br><br>That's what I ended up doing. <br>Also, there's no "obvious" way to index a single file so I ended putting the <br>filename in the glob parameter. <br><br>my $db = Bio::DB::Fasta->new( "$tmp", -glob => "test.faa", -reindex => 1 ); <br><br>--Russell <br><br><div class='shrinkable-quote'><br>> -----Original Message----- <br>> From: Mark A. Jensen [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26507326&i=3" target="_top" rel="nofollow">maj@...</a>] <br>> Sent: Wednesday, 25 November 2009 11:19 a.m. <br>> To: Smithies, Russell; 'bioperl-l' <br>> Subject: Re: [Bioperl-l] Bio::DB::Fasta <br>> <br>> The code (method index_dir() ) seems to expect all the fasta files to be <br>> contained in that directory. Looks hairy; what about creating symlinks to <br>> your <br>> fasta files in a /tmp subdir and calling new() with that subdir? <br>> ----- Original Message ----- <br>> From: "Smithies, Russell" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26507326&i=4" target="_top" rel="nofollow">Russell.Smithies@...</a>> <br>> To: "'bioperl-l'" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26507326&i=5" target="_top" rel="nofollow">bioperl-l@...</a>> <br>> Sent: Tuesday, November 24, 2009 5:04 PM <br>> Subject: [Bioperl-l] Bio::DB::Fasta <br>> <br>> <br>> > Is there any way to pass a filename to Bio::DB::Fasta for the location <br>> of <br>> > where to write the directory.index? <br>> > It's writing in the same dir as the fasta but I'd rather have it write <br>> in /tmp <br>> > as it's part of a web app. <br>> > <br>> > Thanx, <br>> > <br>> > Russell <br>> > <br>> > <br>> > ======================================================================= <br>> > Attention: The information contained in this message and/or attachments <br>> > from AgResearch Limited is intended only for the persons or entities <br>> > to which it is addressed and may contain confidential and/or privileged <br>> > material. Any review, retransmission, dissemination or other use of, or <br>> > taking of any action in reliance upon, this information by persons or <br>> > entities other than the intended recipients is prohibited by AgResearch <br>> > Limited. If you have received this message in error, please notify the <br>> > sender immediately. <br>> > ======================================================================= <br>> > <br>> > _______________________________________________ <br>> > Bioperl-l mailing list <br>> > <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26507326&i=6" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> > <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>> > <br>> > </div><br><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26507326&i=7" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26504898 Remote Blast - same script but different results 2009-11-24T15:00:58Z 2009-11-24T15:00:58Z Chu, Roy Hi bioperl community, <br><br>I've tried searching the old lists to see if this topic has been <br>covered, and perhaps this question arises from my own lack of <br>familiarity with BLAST, but (from my perl script listed below) I get <br>different results with remote blast when I call my script (that is, I <br>will either get hits or no hits at all).  I'll call the script one <br>time, and get no hits.  Then call the script again (with the same <br>parameters), and get the same several hits that I may have before <br>after having gotten no hits.  I use a subroutine to parse the blast <br>report information, and then I use a boolean to indicate whether <br>results are returned or not.  Any insight into what I may have missed <br>would be appreciated.  Short question, is this behavior typical?  My <br>understanding of how BLAST works is that it shouldn'tl... <br><br><br>Thanks in advance, <br>Roy <br><br>#!/usr/bin/perl -w <br><br>use strict; <br>use warnings; <br>use Carp; <br>use Bio::Perl; <br>use CGI; <br>use Bio::SeqIO; <br>use Bio::SearchIO; <br>use Bio::SeqFeature::Generic; <br>use Bio::Restriction::Analysis; <br>use Bio::Tools::Run::RemoteBlast; <br><br>use Bio::SimpleAlign; <br>use Bio::AlignIO; <br>use Bio::LocatableSeq; <br><br>my $five_seqobj = Bio::Seq->new( <br>                -seq => 'ATTCCCACCGGGACCTGCGGGGCTGAGTGCCCTTCTCGGTTGCTGCCGCTGAGGAGCCCGCCCAGCCAGCCAGGGCCGCGAGGCCGAGGCCAGGCCGCAGCCCAGGAGCCGCCCCACCGCAGCTGGCGATGGACCCGCCGAGGCCCGCGCTGCTGGCGCTGCTGGCGCTGCCTGCGCTGCTGCTGCTGCTGCTGGCGGGCGCCAGGGCCG', <br>                -display_id => 'genomic_a', <br>                -alphabet => 'dna', <br>        ); <br>my $three_seqobj = Bio::Seq->new( <br>                -seq => 'GTGAGTGCGCGGCCGCTCTGCGGGCGCAGAGGGAGCGGGAGGGAGCCGGCGGCACGAGGTTGGCCGGGGCAGCCTGGGCCTAGGCCAGAGGGAGGGCAGCCACAGGGTCCAGGGCGAGTGGGGGGATTGGACCAGCTGGCGGCCCCTGCAGGCTCAGGATGGGGGGCGCGGGATGGAGGGGCTGAGGAGGGGGTCTCCGGAGCCTGCCTC', <br>                -display_id => 'genomic_b', <br>                -alphabet => 'dna', <br>        ); <br><br>my @params = ( <br>'-program' => 'blastn', <br>'-database' => 'refseq_genomic', <br>'-expect' => '10', <br>'-readmethod' => 'blastxml' <br>); <br>$Bio::Tools::Run::RemoteBlast::HEADER{'MEGABLAST'} = 'YES'; <br>$Bio::Tools::Run::RemoteBlast::HEADER{'PERC_IDENT'} = 75; <br>$Bio::Tools::Run::RemoteBlast::HEADER{'FORMAT_TYPE'} = 'XML'; <br>$Bio::Tools::Run::RemoteBlast::HEADER{'ENTREZ_QUERY'} = 'Homo sapiens [ORGN]'; <br>$Bio::Tools::Run::RemoteBlast::HEADER{'HITLIST_SIZE'} = 100; # Put: <br>limit number of hits <br><br>my $factory_a = Bio::Tools::Run::RemoteBlast->new(@params); <br>$factory_a->retrieve_parameter('FORMAT_TYPE', 'XML'); <br><br>my $hits_a; <br>my $hits_b; <br><br>my $r; <br>my $bool_hit; <br>print "Submitting BLAST query - 5' end (MEGABLAST = YES)\n"; <br>$Bio::Tools::Run::RemoteBlast::HEADER{'MEGABLAST'} = 'YES'; <br>$r = $factory_a->submit_blast($a_seqobj); <br>$bool_hit = fetch_blast_report($factory_a); <br>unless ($bool_hit) { <br>        print "\nNo hits\n"; <br>        print "Re-submitting BLAST query - 5' end (MEGABLAST = NO)\n"; <br>        sleep 5; <br>        $Bio::Tools::Run::RemoteBlast::HEADER{'MEGABLAST'} = 'NO'; <br>        $r = $factory_a->submit_blast($a_seqobj); <br>        ($bool_hit, $hits_a) = fetch_blast_report($factory_a); <br>        if ($bool_hit == 0) { print "No hits\n"; } <br>        sleep 5; <br>} <br><br>my $factory_b = Bio::Tools::Run::RemoteBlast->new(@params); <br>print "\n--------------------------------------------------\n\n"; <br>print "Submitting BLAST query - 3' end (MEGABLAST = YES)\n"; <br>$Bio::Tools::Run::RemoteBlast::HEADER{'MEGABLAST'} = 'YES'; <br>$r = $remote_blast_three->submit_blast($b_seqobj); <br>$bool_hit = fetch_blast_report($factory_b); <br>unless ($bool_hit) { <br>        print " No hits\n"; <br>        print "Re-submitting BLAST query - 3' end (MEGABLAST = NO)\n"; <br>        sleep 5; <br>        $Bio::Tools::Run::RemoteBlast::HEADER{'MEGABLAST'} = 'NO'; <br>        $r = $factory_b->submit_blast($b_seqobj); <br>        ($bool_hit, $hits_b) = fetch_blast_report($factory_b); <br>        if ($bool_hit == 0) { print " No hits\n"; } <br>        sleep 5; <br>} <br><br>print "\nbye\n\n"; <br><br>print "$hits_a\n$hits_b\n"; <br><br>exit; <br><br>sub fetch_blast_report { <br>        my ($factory) = @_; <br>        my $v = 1; <br>        my $bool_hit = 0; <br>        my $hits = ''; <br>        <br>        print STDERR "waiting..."; <br>        while (my @rids = $factory->each_rid) { <br>                foreach my $rid (@rids) { <br>                        print STDERR "."; <br>                        my $rc = $factory->retrieve_blast($rid); <br>                        # retrieves blast report from remote blast queue, <br>                        # returns -1 on error, 0 on 'job not finished', Bio::SearchIO object <br>                        # args, remote blast id (rid) <br>                        if (!ref($rc)) { <br>                                # if not empty string, ref EXPR returns a non-empty string if EXPR <br>is a reference <br>                                if ($rc < 0) { <br>                                        $factory->remove_rid($rid); <br>                                } <br>                                print STDERR "." if ($v > 0); <br>##################################################################################### <br>is this printing out as multiple dots? when and why? <br>                                sleep 5; <br>                        } else { <br>                                $bool_hit = 1; <br>                                my $result = $rc->next_result(); <br>                                unless ($result->num_hits > 0) { <br>                                        $bool_hit = 0; <br>                                } <br>                                # returns: Bio::Search::Result::ResultI object <br>                                $factory->remove_rid($rid); <br>                                print "\ndatabase:\t", $result->database_name,"\n"; <br>                                print "query name:\t", $result->query_name,"\n"; <br>                                print "query length\t", $result->query_length,"\n"; <br>                                print "num hits\t", $result->num_hits,"\n"; <br>                                if ($result->num_hits) { <br>                                        # $result->hits returns an array of hits <br>                                        # $results->no_hits_found, boolean vs $#{@hits} ie. filtering\ <br>                                        while (my $hit = $result->next_hit) { <br>                                        <br>                                        print "\nhit name:\t", $hit->name,"\n"; <br>                                        print "description:\t", $hit->description,"\n"; <br>                                        print "locus:\t", $hit->locus,"\n"; <br>                                        print "algorithm: ", $hit->algorithm,"\thit length: ", <br>$hit->length,"\thit ranking: ", $hit->rank,"\n"; <br>                                        while (my $hsp = $hit->next_hsp) { <br>                                                print "evalue: ", $hsp->evalue,"\tscore: ", <br>$hsp->score,"\tpercent_id: ", $hsp->percent_identity,"\n"; <br>                                                print "query_start: ", $hsp->query->start,"\tquery_end: ", <br>$hsp->query->end; <br>                                                print "\tquery_length: ", $hsp->query->length,"\tquery_strand: <br>", $hsp->strand('query'), "\n"; <br>                                                print "subject_start: ", $hsp->subject->start,"\tsubject_end: ", <br>$hsp->subject->end; <br>                                                print "\tsubject_length: ", <br>$hsp->subject->length,"\tsubject_strand: ", $hsp->strand('subject'), <br>"\n\n"; <br>                                                my $aln = $hsp->get_aln; <br>                                                if ($aln->is_flush) { <br>                                                        foreach my $seq ($aln->each_seq) { <br>                                                                print $seq->seq,"\n"; <br>                                                        } <br>                                                        print $aln->gap_line, "\n"; <br>                                                        print $aln->consensus_string(95), "\n\n"; <br>                                                } <br><br>                                                $hits .= $hit->name."\t".$hsp->subject->start."\t".$hsp->subject->end."\t".$hsp->strand('subject')."\n"; <br>                                        } <br>                                } <br>                        } <br>                } <br>        } <br>        return ($bool_hit, $hits); <br>} <br>} <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504898&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26504793 Re: Bio::Tools::Run::Cap3 usage question 2009-11-24T14:54:48Z 2009-11-24T14:54:48Z Florent Angly Hi Paolo, <br><br>It turns out that there is no standard for what is to be passed to the <br>Bio::Tools::Run wrappers and returned by them. I noticed the <br>inconsistency between the assembly wrappers recently while implementing <br>support for new wrapper. I implemented inital support for additional de <br>novo assembly programs in BioPerl (454 Newbler and Minimo) a couple of <br>weeks ago and Mark Jensen added support for Maq, a program that <br>assembler reads against a reference. In the process, all the assembly <br>wrappers were changed to take the same type of input data (a FASTA <br>sequence or an array reference of sequence objects) and return one of <br>the following: <br>    * a Bio::Assembly::Scaffold object (the default), or <br>    * a Bio::Assembly::IO object, or <br>    * the name of a file for the output of the assembler <br>Use the out_type method to set up which output you want, e.g.: <br>    $factory->out_type('Bio::Assembly::IO'); <br>or <br>    $factory->out_type('cap3_results.ace'); <br>You'll have to use the code in the bioperl-run subversion if you want to <br>use these new features. <br><br>Cheers, <br><br>Florent <br><br><br><br><br>Paolo Pavan wrote: <div class='shrinkable-quote'><br>> Dear, <br>> I'm confused about the proper usage of the module Bio::Tools::Run::Cap3. <br>> As documented in the pod, the run(@seqs) method returns the cap3 report file <br>> while I expect to return a Bio::Assembly object, consistently with other <br>> Bio::Tools::Run classes. <br>> However, I went around this by getting from the factory object the location <br>> and the names of the temp output files (actually accessing a private <br>> property, although) and reading them via the Assembly::IO system. <br>> I was just wandering what is the proper designed way to do this job. <br>> <br>> Thank you for enlighten the way! <br>> Paolo <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504793&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>> <br>>   </div><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504793&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26504347 Re: Bio::DB::Fasta 2009-11-24T14:21:52Z 2009-11-24T14:21:52Z Smithies, Russell That's what I ended up doing. <br>Also, there's no "obvious" way to index a single file so I ended putting the filename in the glob parameter. <br><br>my $db = Bio::DB::Fasta->new( "$tmp", -glob => "test.faa", -reindex => 1 ); <br><br>--Russell <br><br><div class='shrinkable-quote'><br>> -----Original Message----- <br>> From: Mark A. Jensen [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504347&i=0" target="_top" rel="nofollow">maj@...</a>] <br>> Sent: Wednesday, 25 November 2009 11:19 a.m. <br>> To: Smithies, Russell; 'bioperl-l' <br>> Subject: Re: [Bioperl-l] Bio::DB::Fasta <br>> <br>> The code (method index_dir() ) seems to expect all the fasta files to be <br>> contained in that directory. Looks hairy; what about creating symlinks to <br>> your <br>> fasta files in a /tmp subdir and calling new() with that subdir? <br>> ----- Original Message ----- <br>> From: "Smithies, Russell" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504347&i=1" target="_top" rel="nofollow">Russell.Smithies@...</a>> <br>> To: "'bioperl-l'" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504347&i=2" target="_top" rel="nofollow">bioperl-l@...</a>> <br>> Sent: Tuesday, November 24, 2009 5:04 PM <br>> Subject: [Bioperl-l] Bio::DB::Fasta <br>> <br>> <br>> > Is there any way to pass a filename to Bio::DB::Fasta for the location <br>> of <br>> > where to write the directory.index? <br>> > It's writing in the same dir as the fasta but I'd rather have it write <br>> in /tmp <br>> > as it's part of a web app. <br>> > <br>> > Thanx, <br>> > <br>> > Russell <br>> > <br>> > <br>> > ======================================================================= <br>> > Attention: The information contained in this message and/or attachments <br>> > from AgResearch Limited is intended only for the persons or entities <br>> > to which it is addressed and may contain confidential and/or privileged <br>> > material. Any review, retransmission, dissemination or other use of, or <br>> > taking of any action in reliance upon, this information by persons or <br>> > entities other than the intended recipients is prohibited by AgResearch <br>> > Limited. If you have received this message in error, please notify the <br>> > sender immediately. <br>> > ======================================================================= <br>> > <br>> > _______________________________________________ <br>> > Bioperl-l mailing list <br>> > <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504347&i=3" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> > <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>> > <br>> > </div><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504347&i=4" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26504502 Re: Bio::DB::Fasta 2009-11-24T14:18:51Z 2009-11-24T14:18:51Z Mark A. Jensen The code (method index_dir() ) seems to expect all the fasta files to be <br>contained in that directory. Looks hairy; what about creating symlinks to your <br>fasta files in a /tmp subdir and calling new() with that subdir? <br>----- Original Message ----- <br>From: "Smithies, Russell" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504502&i=0" target="_top" rel="nofollow">Russell.Smithies@...</a>> <br>To: "'bioperl-l'" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504502&i=1" target="_top" rel="nofollow">bioperl-l@...</a>> <br>Sent: Tuesday, November 24, 2009 5:04 PM <br>Subject: [Bioperl-l] Bio::DB::Fasta <br><br><div class='shrinkable-quote'><br>> Is there any way to pass a filename to Bio::DB::Fasta for the location of <br>> where to write the directory.index? <br>> It's writing in the same dir as the fasta but I'd rather have it write in /tmp <br>> as it's part of a web app. <br>> <br>> Thanx, <br>> <br>> Russell <br>> <br>> <br>> ======================================================================= <br>> Attention: The information contained in this message and/or attachments <br>> from AgResearch Limited is intended only for the persons or entities <br>> to which it is addressed and may contain confidential and/or privileged <br>> material. Any review, retransmission, dissemination or other use of, or <br>> taking of any action in reliance upon, this information by persons or <br>> entities other than the intended recipients is prohibited by AgResearch <br>> Limited. If you have received this message in error, please notify the <br>> sender immediately. <br>> ======================================================================= <br>> <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504502&i=2" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>> <br>> </div><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504502&i=3" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26504203 Bio::DB::Fasta 2009-11-24T14:04:31Z 2009-11-24T14:04:31Z Smithies, Russell Is there any way to pass a filename to Bio::DB::Fasta for the location of where to write the directory.index? <br>It's writing in the same dir as the fasta but I'd rather have it write in /tmp as it's part of a web app. <br><br>Thanx, <br><br>Russell <br><br><br>======================================================================= <br>Attention: The information contained in this message and/or attachments <br>from AgResearch Limited is intended only for the persons or entities <br>to which it is addressed and may contain confidential and/or privileged <br>material. Any review, retransmission, dissemination or other use of, or <br>taking of any action in reliance upon, this information by persons or <br>entities other than the intended recipients is prohibited by AgResearch <br>Limited. If you have received this message in error, please notify the <br>sender immediately. <br>======================================================================= <br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26504203&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26501967 Bio::Tools::Run::Cap3 usage question 2009-11-24T11:28:09Z 2009-11-24T11:28:09Z Paolo Pavan-2 Dear, <br>I'm confused about the proper usage of the module Bio::Tools::Run::Cap3. <br>As documented in the pod, the run(@seqs) method returns the cap3 report file <br>while I expect to return a Bio::Assembly object, consistently with other <br>Bio::Tools::Run classes. <br>However, I went around this by getting from the factory object the location <br>and the names of the temp output files (actually accessing a private <br>property, although) and reading them via the Assembly::IO system. <br>I was just wandering what is the proper designed way to do this job. <br><br>Thank you for enlighten the way! <br>Paolo <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26501967&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26499002 Re: verbosity and error stack, was accessing EMBL database 2009-11-24T08:18:22Z 2009-11-24T08:18:22Z Dave Messina-3 > I like the idea of optionally having a log class, if someone comes up with a nice way of adding it in I would be for it. <br><br>My suggestion of the logging modules was actually to handle the various levels of verbose output -- I think both of the ones I mentioned "log" to STDERR by default. <br><br>But of course a nice side effect of using such a logging module is that it would allow optional logging to a file, too. <br><br>Dave <br><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26499002&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26490011 Re: verbosity and error stack, was accessing EMBL database 2009-11-23T19:19:02Z 2009-11-23T19:19:02Z Chris Fields-5 Okay, so I think it's feasible to add this into trunk.  I like the idea of optionally having a log class, if someone comes up with a nice way of adding it in I would be for it. <br><br>chris <br><br>On Nov 20, 2009, at 4:15 AM, Dave Messina wrote: <br><div class='shrinkable-quote'><br>> Chris, I took a look at how you implemented this in Biome -- very nice! <br>> <br>> <br>>> I like this verbose/strict separability a lot. Should we go for it? <br>> <br>> Me too. So yes, I think so. <br>> <br>> <br>>> We could even allow finer-grained control of verbosity (states which cover all combinations) w/o affecting strictness. <br>> <br>> <br>> Perhaps this is a job for Log::Log4Perl or Log::Dispatch? <br>> <a href="http://search.cpan.org/~mschilli/Log-Log4perl-1.25/lib/Log/Log4perl.pm" target="_top" rel="nofollow">http://search.cpan.org/~mschilli/Log-Log4perl-1.25/lib/Log/Log4perl.pm</a><br>> <a href="http://search.cpan.org/~drolsky/Log-Dispatch-2.26/lib/Log/Dispatch.pm" target="_top" rel="nofollow">http://search.cpan.org/~drolsky/Log-Dispatch-2.26/lib/Log/Dispatch.pm</a><br>> <br>> <br>> That might be overkill, though. <br>> <br>> Dave <br>> </div><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26490011&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26484217 Re: Installing Bio-perl on dreamhost via CPAN 2009-11-23T11:13:26Z 2009-11-23T11:13:26Z Jonathan Cline Dreamhost has terrible reliability.  I have stats going back years on a <br>standard dreamhost hosting account (non-dedicated server), and on some <br>days the web server doesn't respond.  Dreamhost service is OK for a <br>hobby blog however it is definitely *not* suitable for anything real. <br>Add in latency, arbitrary account limits/restrictions,  etc, and as a <br>hosting service, it is a bad idea to host a project there.   Although <br>some users apparently get lucky with server allocation and end up on a <br>"good server", the provider can change this at any time as well.  I <br>think more typically, the accounts users don't notice, since most are <br>simple bloggers. <br><br>Here's a data snip that illustrates the problem with a typical dreamhost <br>account: <br><br>---------------------------------------------------------------------- <br>date          uptime       dns   connect   request      ttfb      ttlb <br><br>2008-08-05     91.40     0.000     0.528     0.528     2.257     1.619 <br>2008-08-04     89.13     0.002     0.301     0.301     1.302     0.971 <br>2008-08-03     94.62     0.000     0.567     0.567     1.506     0.913 <br>2008-08-02    100.00     0.000     0.335     0.335     1.475     1.079 <br>2008-08-01    100.00     0.000     0.310     0.310     1.587     0.825 <br>2008-07-31     93.55     0.023     0.386     0.386     1.280     0.759 <br>2008-07-30    100.00     0.000     0.345     0.345     1.373     0.860 <br>2008-07-29    100.00     0.000     0.358     0.358     1.335     0.757 <br>2008-07-28    100.00     0.000     0.327     0.327     1.462     0.896 <br>2008-07-27    100.00     0.000     0.292     0.292     1.410     0.966 <br>2008-07-26    100.00     0.000     0.283     0.283     1.280     0.815 <br>2008-07-25    100.00     0.000     0.297     0.297     1.231     0.853 <br>2008-07-24    100.00     0.000     0.362     0.362     1.258     0.699 <br>2008-07-23    100.00     0.000     0.339     0.339     1.270     0.785 <br><br>---------------------------------------------------------------------- <br>minimum        89.13     0.000     0.283     0.283     1.231     0.699 <br>maximum       100.00     0.023     0.567     0.567     2.257     1.619 <br>average        97.76     0.002     0.359     0.359     1.430     0.914 <br>---------------------------------------------------------------------- <br><br><br>Or this month: <br><br>---------------------------------------------------------------------- <br>date          uptime       dns   connect   request      ttfb      ttlb <br><br>2009-11-11    100.00     0.011     0.097     0.097     1.260     1.638 <br>2009-11-10    100.00     0.008     0.094     0.094     1.285     1.647 <br>2009-11-09    100.00     0.008     0.094     0.094     1.494     1.872 <br>2009-11-08    100.00     0.015     0.101     0.101     1.509     1.894 <br>2009-11-07    100.00     0.006     0.092     0.092     1.453     1.831 <br>2009-11-06    100.00     0.011     0.097     0.097     1.500     1.882 <br>2009-11-05     97.80     0.012     0.097     0.097     1.445     1.806 <br>2009-11-04    100.00     0.010     0.096     0.096     1.235     1.605 <br>2009-11-03     95.65     0.007     0.093     0.093     1.266     1.612 <br>2009-11-02    100.00     0.010     0.096     0.096     1.267     1.637 <br>2009-11-01    100.00     0.007     0.093     0.093     1.311     1.692 <br>2009-10-31    100.00     0.009     0.095     0.095     1.225     1.594 <br>2009-10-30    100.00     0.009     0.095     0.095     1.364     1.739 <br>2009-10-29    100.00     0.017     0.103     0.103     1.121     1.505 <br><br>---------------------------------------------------------------------- <br>minimum        95.65     0.006     0.092     0.092     1.121     1.505 <br>maximum       100.00     0.017     0.103     0.103     1.509     1.894 <br>average        99.53     0.010     0.096     0.096     1.338     1.711 <br>---------------------------------------------------------------------- <br><br><br>## Jonathan Cline <br>## <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26484217&i=0" target="_top" rel="nofollow">jcline@...</a> <br>## Mobile: +1-805-617-0223 <br>######################## <br><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26484217&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26483454 building samtools/Bio::DB::Sam on cygwin 2009-11-23T09:58:46Z 2009-11-23T09:58:46Z Mark A. Jensen Hi All-- <br><br>I've had some hard-won success installing samtools and Lincoln's <br>Bio::DB::Sam under cygwin; thought some on the list would be able to <br>use my notes. (Yes, Jason, I'm working on Bio::Tools::Run::BWA...) <br><br><br>(To get the current samtools, ping <br><a href="http://sourceforge.net/projects/samtools/files/samtools/0.1.7/samtools-0.1.7a.tar.bz2/download" target="_top" rel="nofollow">http://sourceforge.net/projects/samtools/files/samtools/0.1.7/samtools-0.1.7a.tar.bz2/download</a><br>) <br><br>* Getting samtools to make from scratch in cygwin <br><br>The following diff details the changes to the samtools Makefile I made <br>by hand. The key points are <br><br>-D_WIN32 <br><br>and the additional variable LFLAGS and its interpolations. To get the <br>linker to see <br><br>libgcc libstdc++ <br><br>I needed to add symlinks from /lib to the correct files in <br>/lib/gcc/i386-pc-cygwin/4.3.2/. Your gcc version may differ. <br><br><br>--- ../old/samtools-0.1.7a/Makefile 2009-11-16 10:13:43.000000000 -0500 <br>+++ Makefile 2009-11-23 12:14:18.529000000 -0500 <br>@@ -1,16 +1,18 @@ <br> CC=   gcc <br> CFLAGS=  -g -Wall -O2 #-m64 #-arch ppc <br>-DFLAGS=  -D_FILE_OFFSET_BITS=64 -D_USE_KNETFILE -D_CURSES_LIB=1 <br>+LFLAGS=         -lws2_32 -lgcc -lcygwin -lbz2 -lz -lstdc++ <br>+DFLAGS=  -D_WIN32 -D_FILE_OFFSET_BITS=64 -D_CURSES_LIB=1 <br> LOBJS=  bgzf.o kstring.o bam_aux.o bam.o bam_import.o sam.o bam_index.o \ <br>    bam_pileup.o bam_lpileup.o bam_md.o glf.o razf.o faidx.o knetfile.o \ <br>    bam_sort.o sam_header.o <br> AOBJS=  bam_tview.o bam_maqcns.o bam_plcmd.o sam_view.o \ <br>    bam_rmdup.o bam_rmdupse.o bam_mate.o bam_stat.o bam_color.o \ <br>    bamtk.o kaln.o <br><br>@@ -36,13 +38,13 @@ <br>   $(AR) -cru $@ $(LOBJS) <br>  <br> samtools:lib $(AOBJS) <br>-  $(CC) $(CFLAGS) -o $@ $(AOBJS) -lm $(LIBPATH) $(LIBCURSES) -lz -L. -lbam <br>+  $(CC) $(CFLAGS) -o $@ $(AOBJS) -Xlinker --enable-auto-import -lm $(LIBPATH) $(LIBCURSES) -lz -L. -lbam $(LFLAGS) <br>  <br> razip:razip.o razf.o knetfile.o <br>-  $(CC) $(CFLAGS) -o $@ razf.o razip.o knetfile.o -lz <br>+  $(CC) $(CFLAGS) -o $@ razf.o razip.o knetfile.o -lz -lm -lws2_32 <br>  <br> bgzip:bgzip.o bgzf.o <br>-  $(CC) $(CFLAGS) -o $@ bgzf.o bgzip.o -lz <br>+  $(CC) $(CFLAGS) -o $@ bgzf.o bgzip.o -lz -lm -lws2_32 <br>  <br> razip.o:razf.h <br> bam.o:bam.h razf.h bam_endian.h kstring.h sam_header.h <br><br>* Getting Bio::DB::Sam to compile and install <br><br>Bio::DB::Sam requires not the samtools.exe, but the bam library <br>created during the samtools build, as well as all the samtools header <br>files. Create a symlink in /lib to libbam.a in the build directory (or <br>copy libbam.a up to /lib), and create symlinks or copy *.h into <br>/usr/include. Then in cygwin bash shell <br><br>$ cpan <br>cpan> install Bio::DB::Sam <br><br>should fly. <br><br>Hope someone finds this useful. These mods led me to a successful <br>Bio::DB::Sam install--have not yet checked original code based on <br>Bio::DB::Sam. If they don't work for you, reply to the list. <br><br>cheers, <br>MAJ <br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26483454&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26482702 Re: Problems with Genbank Proteins File 2009-11-23T09:36:36Z 2009-11-23T09:36:36Z Peter-329 On Sun, Nov 22, 2009 at 4:06 PM, Rodrigo Jardim <br><<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26482702&i=0" target="_top" rel="nofollow">jardim.rodrigo@...</a>> wrote: <br>> I have been problem to parser genbank protein file. I think that because <br>> this file have a other order of fields. For example: <br>> <br>> ... <br>> <br>> When I try to parser, Bioperl abort with message error. <br>> <br>> Any ideas? <br><br>There are some important bits of information missing - what is the error <br>message, and what version of BioPerl are you using? <br><br>Peter <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26482702&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26469344 Re: BioPerl on FLOSS Weekly 2009-11-22T12:48:39Z 2009-11-22T12:48:39Z Mark A. Jensen We do have <a href="http://www.bioperl.org/wiki/Category:BioPerl_Media" target="_top" rel="nofollow">http://www.bioperl.org/wiki/Category:BioPerl_Media</a> -- <br>----- Original Message ----- <br>From: "Chris Fields" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26469344&i=0" target="_top" rel="nofollow">cjfields@...</a>> <br>To: "Adam Sjøgren" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26469344&i=1" target="_top" rel="nofollow">adsj@...</a>> <br>Cc: <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26469344&i=2" target="_top" rel="nofollow">bioperl-l@...</a>> <br>Sent: Sunday, November 22, 2009 3:30 PM <br>Subject: Re: [Bioperl-l] BioPerl on FLOSS Weekly <br><br><br><br>On Nov 22, 2009, at 6:37 AM, Adam Sjøgren wrote: <br><div class='shrinkable-quote'><br>> On Sat, 21 Nov 2009 20:58:37 -0600, Chris wrote: <br>> <br>>> Jason and I were recently interviewed (Wednesday!) about BioPerl for <br>>> FLOSS Weekly by Randal Schwartz, Leo Laporte, Marc Pelletier, and <br>>> Kirsten Sanford. <br>> <br>> Great! <br>> <br>> How about linking to it on bioperl.org? <br>> <br>> <br>>  :-), <br>> <br>>   Adam <br>> <br>> -- <br>>                                                          Adam Sjøgren <br>>                                                    <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26469344&i=3" target="_top" rel="nofollow">adsj@...</a> </div><br>Now posted via O|B|F News; I'll try to make that feed more prominent on the main <br>page. <br><br>Since this is the second such interview (Jason did one a few years back for <br>PerlCast), I'm thinking we need a media page of some sort. <br><br>chris <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26469344&i=4" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26469344&i=5" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26469276 Re: BioPerl on FLOSS Weekly 2009-11-22T12:30:01Z 2009-11-22T12:30:01Z Chris Fields-5 <br>On Nov 22, 2009, at 6:37 AM, Adam Sjøgren wrote: <br><div class='shrinkable-quote'><br>> On Sat, 21 Nov 2009 20:58:37 -0600, Chris wrote: <br>> <br>>> Jason and I were recently interviewed (Wednesday!) about BioPerl for <br>>> FLOSS Weekly by Randal Schwartz, Leo Laporte, Marc Pelletier, and <br>>> Kirsten Sanford. <br>> <br>> Great! <br>> <br>> How about linking to it on bioperl.org? <br>> <br>> <br>>  :-), <br>> <br>>   Adam <br>> <br>> -- <br>>                                                          Adam Sjøgren <br>>                                                    <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26469276&i=0" target="_top" rel="nofollow">adsj@...</a> </div><br>Now posted via O|B|F News; I'll try to make that feed more prominent on the main page.   <br><br>Since this is the second such interview (Jason did one a few years back for PerlCast), I'm thinking we need a media page of some sort. <br><br>chris <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26469276&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26482406 Problems with Genbank Proteins File 2009-11-22T08:06:40Z 2009-11-22T08:06:40Z Rodrigo Jardim-3 I have been problem to parser genbank protein file. I think that because <br>this file have a other order of fields. For example: <br><br>In most general genbank files: <br>======================== <br>LOCUS       AA399704                  183 bp   mRNA    linear   EST <br>03-MAR-2000 <br>ACCESSION   AA399704 <br>VERSION     AA399704.1  GI:2053305 <br>DEFINITION  TEUF0001 T.cruzi epimastigote non-normalized cDNA Library <br>            Trypanosoma cruzi cDNA clone 1 5' similar to T. cruzi gene for <br>            histone H2b (X60982), mRNA sequence. <br>KEYWORDS    EST. <br>SOURCE      Trypanosoma cruzi <br><br>In genbank protein files: <br>=================== <br>LOCUS       XP_628849                510 aa            linear   INV <br>31-OCT-2008 <br>DEFINITION  hypothetical protein [Dictyostelium discoideum AX4]. <br>ACCESSION   XP_628849 <br>VERSION     XP_628849.1  GI:66799847 <br>DBSOURCE    REFSEQ: accession XM_628847.1 <br>KEYWORDS    . <br>SOURCE      Dictyostelium discoideum AX4. <br><br>When I try to parser, Bioperl abort with message error. <br><br>Any ideas? <br><br>Thanks all, <br><br>-- <br>Atc, <br>Rodrigo Jardim <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26482406&i=0" target="_top" rel="nofollow">jardim.rodrigo@...</a> <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26482406&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26465032 Re: BioPerl on FLOSS Weekly 2009-11-22T04:37:40Z 2009-11-22T04:37:40Z Adam Sjøgren-2 On Sat, 21 Nov 2009 20:58:37 -0600, Chris wrote: <br><br>> Jason and I were recently interviewed (Wednesday!) about BioPerl for <br>> FLOSS Weekly by Randal Schwartz, Leo Laporte, Marc Pelletier, and <br>> Kirsten Sanford. <br><br>Great! <br><br>How about linking to it on bioperl.org? <br><br><br>  :-), <br><br>   Adam <br><br>-- <br>                                                          Adam Sjøgren <br>                                                    <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26465032&i=0" target="_top" rel="nofollow">adsj@...</a> <br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26465032&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26462491 BioPerl on FLOSS Weekly 2009-11-21T18:58:37Z 2009-11-21T18:58:37Z Chris Fields-5 Jason and I were recently interviewed (Wednesday!) about BioPerl for FLOSS Weekly by Randal Schwartz, Leo Laporte, Marc Pelletier, and Kirsten Sanford.  The interview is now available online, so get your favorite flavor (MP3, podcast) here: <br><br><a href="http://twit.tv/floss96" target="_top" rel="nofollow">http://twit.tv/floss96</a><br><br>Enjoy! <br><br>chris and jason <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26462491&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26459489 Re: Bio::DB::EUtilities question 2009-11-21T11:19:24Z 2009-11-21T11:19:24Z Chris Fields-5 NCBI has specific rules about the repeated queries to its servers: <br><br><a href="http://eutils.ncbi.nlm.nih.gov/#UserSystemRequirements" target="_top" rel="nofollow">http://eutils.ncbi.nlm.nih.gov/#UserSystemRequirements</a><br><br>Acc. to that, if you are making over 100 requests at peak times you will run into problems (they'll probably temp-block your IP), even if the timeout is much shorter now (it's 3 requests/second, whereas a year or two ago it was once every 3 sec).  In general it's best to run something like this during off-hours.   <br><br>The actual limit on number of server requests is one specific part of Bio::DB::EUtilities that hasn't been added yet, but is tentatively planned.   <br><br>chris <br><br>On Nov 21, 2009, at 12:26 PM, Robert Bradbury wrote: <br><div class='shrinkable-quote'><br>> It sounds like NCBI may be counting frequency of requests, how much data <br>> they send or something similar.  Are you delaying the time between fetches? <br>> The code I've seen typically sleeps for a few seconds each time around a <br>> loop.  You might try longer delays between fetches and see if that gets you <br>> any more data. <br>> <br>> Alternatively perhaps the libraries aren't reusing the TCP/IP connection <br>> properly.  Is there a difference between the amount of memory on the <br>> machines?  Have you watched the size of the process to see if it grows over <br>> time?  I think the bug which prevented me from fetching a not-so-large <br>> genome from a few months ago (eating up 3GB of memory in the process) has <br>> not been resolved.  If so that could be your problem. <br>> <br>> Robert <br>> <br>> On Fri, Nov 20, 2009 at 12:44 PM, Alessandra <br>> <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26459489&i=0" target="_top" rel="nofollow">alessandra.bilardi@...</a>>wrote: <br>>> <br>>> <br>>> I'm testing Bio::DB::EUtilities - webagent which interacts with and <br>>> retrieves data from NCBI's eUtils. My perl script works but it works <br>>> only if I request less than ~450 times get_Response function.. else I <br>>> have got this error message: <br>>> <br>>> ------------- EXCEPTION ------------- <br>>> MSG: Response Error <br>>> Can't connect to eutils.ncbi.nlm.nih.gov:80 (connect: No route to host) <br>>> STACK Bio::DB::GenericWebAgent::get_Response <br>>> /usr/local/share/perl/5.10.0/Bio/DB/GenericWebAgent.pm:215 <br>>> STACK toplevel ./wget4gbk.pl:77 <br>>> <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26459489&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a></div><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26459489&i=2" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26459097 Re: Bio::DB::EUtilities question 2009-11-21T10:26:05Z 2009-11-21T10:26:05Z RobertBradbury It sounds like NCBI may be counting frequency of requests, how much data <br>they send or something similar.  Are you delaying the time between fetches? <br> The code I've seen typically sleeps for a few seconds each time around a <br>loop.  You might try longer delays between fetches and see if that gets you <br>any more data. <br><br>Alternatively perhaps the libraries aren't reusing the TCP/IP connection <br>properly.  Is there a difference between the amount of memory on the <br>machines?  Have you watched the size of the process to see if it grows over <br>time?  I think the bug which prevented me from fetching a not-so-large <br>genome from a few months ago (eating up 3GB of memory in the process) has <br>not been resolved.  If so that could be your problem. <br><br>Robert <br><br>On Fri, Nov 20, 2009 at 12:44 PM, Alessandra <br><<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26459097&i=0" target="_top" rel="nofollow">alessandra.bilardi@...</a>>wrote: <div class='shrinkable-quote'><br>> <br>> <br>> I'm testing Bio::DB::EUtilities - webagent which interacts with and <br>> retrieves data from NCBI's eUtils. My perl script works but it works <br>> only if I request less than ~450 times get_Response function.. else I <br>> have got this error message: <br>> <br>> ------------- EXCEPTION ------------- <br>> MSG: Response Error <br>> Can't connect to eutils.ncbi.nlm.nih.gov:80 (connect: No route to host) <br>> STACK Bio::DB::GenericWebAgent::get_Response <br>> /usr/local/share/perl/5.10.0/Bio/DB/GenericWebAgent.pm:215 <br>> STACK toplevel ./wget4gbk.pl:77 <br>> </div>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26459097&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26453982 Re: Excessive CPU use by various Bioperl sites 2009-11-20T21:20:37Z 2009-11-20T21:20:37Z Chris Fields-5 On Nov 20, 2009, at 11:11 PM, Sean Davis wrote: <br><div class='shrinkable-quote'><br>> On Fri, Nov 20, 2009 at 8:38 PM, Chris Fields <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453982&i=0" target="_top" rel="nofollow">cjfields@...</a>> wrote: <br>> <br>>> Robert, <br>>> <br>>> Not sure why you're seeing that, but the HOWTO (and, AFAIK, the wiki in <br>>> general) do not use JS, unless there is a specific addition I'm unaware of. <br>>> Now, the site wiki was recently 'parasited' for redirects, which may be the <br>>> culprit, but this is now fixed.  Can you at least retest to see if this <br>>> persists? <br>>> <br>>> Anyone else know about this? <br>>> <br>>> <br>> The page in question does include javascript, it appears from the source. <br>> This is a function of using mediawiki, though, I believe and not something <br>> specific to that page. <br>> <br>> Sean </div><br></sound of my hand slapping my forehead> <br><br>Sean, thanks for pointing that out. <br><br>chris <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453982&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26453926 Re: Excessive CPU use by various Bioperl sites 2009-11-20T21:11:34Z 2009-11-20T21:11:34Z Sean Davis-3 On Fri, Nov 20, 2009 at 8:38 PM, Chris Fields <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453926&i=0" target="_top" rel="nofollow">cjfields@...</a>> wrote: <br><div class='shrinkable-quote'><br>> Robert, <br>> <br>> Not sure why you're seeing that, but the HOWTO (and, AFAIK, the wiki in <br>> general) do not use JS, unless there is a specific addition I'm unaware of. <br>>  Now, the site wiki was recently 'parasited' for redirects, which may be the <br>> culprit, but this is now fixed.  Can you at least retest to see if this <br>> persists? <br>> <br>> Anyone else know about this? <br>> <br>> </div>The page in question does include javascript, it appears from the source. <br> This is a function of using mediawiki, though, I believe and not something <br>specific to that page. <br><br>Sean <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453926&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26453814 Re: Excessive CPU use by various Bioperl sites 2009-11-20T20:38:23Z 2009-11-20T20:38:23Z Chris Fields-5 Robert, <br><br>Not sure why you're seeing that, but the HOWTO (and, AFAIK, the wiki in general) do not use JS, unless there is a specific addition I'm unaware of.  Now, the site wiki was recently 'parasited' for redirects, which may be the culprit, but this is now fixed.  Can you at least retest to see if this persists? <br><br>Anyone else know about this? <br><br>chris <br><br>On Nov 20, 2009, at 7:40 PM, Robert Bradbury wrote: <br><div class='shrinkable-quote'><br>> I run a Linux system which is in a gradual process of evolution from the <br>> default Linux browsers (Galeon, Epiphany, etc.) through Firefox (better) to <br>> Google's Chromium (IMO, perhaps the best so far).  Chromium allows one to <br>> create a process per tab/URL so one can effectively track what it is doing. <br>> It also allows one to track the machine usage of these processes (through <br>> the Developer > Task manager [shift-escape keyboard] option) which though <br>> expensive in terms of overhead allows one to track offending windows (in <br>> terms of memory or CPU use).  My processor recently jumped from a typical <br>> 700 MHz to 1.4 GHz speed (using the Linux Ondemand scheduler - which saves <br>> ~20 W at the wall outlet -- I've measured it) to the full tilt 2.8 GHz the <br>> CPU is capable of.  Looking at the chrome task manager I was not surprised <br>> to find the NY Times high on the list (they are pushing content, esp. using <br>> Javascript) but much to my dismay the Jalview and Howto:Trees:Bioperl <br>> appeared to be high on the list.  Now I am forced to ask myself *why* sites <br>> which are simply distributing static information are eating up CPU on my <br>> machine!  This is a fundamental flaw in the architecture of the sites -- <br>> wherein there should be conscious efforts to minimize user-CPU use (or avoid <br>> Javascript entirely).  This would not be a problem if I were using Firefox <br>> as I can easily use NoScript to block Javacscript from non-approved sites. <br>> But it raises the question of when one should allow Javascript to run (one <br>> would "normally" approve academic sites by default) when even the academic <br>> sites are abusing my CPU.  There needs to be much greater awareness both on <br>> the part of software distributors and software consumers that it is *MY* CPU <br>> and *MY* Electricty and *MY* contribution to global warming.  And the <br>> developers/distributors should not be sucking down those resources without <br>> first saying "May I?" and I have the option of saying "No you may not." <br>> There is enough we can do productively (running low homology blast <br>> searches) without engaging in endless wheel spinning of Javascripts or <br>> looped GIFs. <br>> <br>> Robert <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453814&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a></div><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453814&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26453764 Re: Installing Bio-perl on dreamhost via CPAN 2009-11-20T20:33:02Z 2009-11-20T20:33:02Z Mark A. Jensen Maybe 'nightmarehost' is more appropriate. I've had no problems on AWS, <br>but this may not exactly what you need. MAJ <br>----- Original Message ----- <br>From: "Chu, Roy" <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453764&i=0" target="_top" rel="nofollow">roychu@...</a>> <br>To: <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453764&i=1" target="_top" rel="nofollow">bioperl-l@...</a>> <br>Sent: Friday, November 20, 2009 3:23 PM <br>Subject: Re: [Bioperl-l] Installing Bio-perl on dreamhost via CPAN <br><br><br>"sounds very much like you process was killed for prolonged execution <br>time, or memory usage. We have a daemon in place that monitors for <br>processes that take up too much of a shared web server's resources, and <br>this may have kicked in (and often does when trying to install packages <br>on a shared server)." <br><br>This was the explanation they had.  Regarding asking their admins to <br>install, it seems is a "they'll try to get to it but don't hold your <br>breath situation." <br><br>Hmmm, I tried some other attempts, installing 1.4.0 posed no problems. <br> I'm not a perl guru, so I tried to increase the build cache size from <br>the default, 10 MB, hoping that that may be the problem--can't imagine <br>how though, since I can't imagine how big the whole package version <br>can differ by (though honestly, I haven't checked). <br>Whenever I try to install 1.6.1, it runs into a problem I guess after <br>the 'make' step and lists the <br>modules--BioPerl-1.6.0/t/Variation/SeqDiff.t <br>BioPerl-1.6.0/t/Variation/SNP.t <br>BioPerl-1.6.0/t/Variation/Variation_IO.t <br>--and typically gets killed here '> Killed' <br><br>Next, I tried 1.6.0, then I get this: <br>"(I think you ran Build.PL directly, so will use CPAN to install <br>prerequisites on demand) <br>CPAN: Storable loaded ok (v2.12) <br>Going to read '/home/$username/.cpan/Metadata' <br>Killed" (everything prior works and it seems to get further along than <br>when I try to install 1.6.1) <br><br>Any insight into why this may be happening would be appreciated. <br>Something EQUALLY appreciated would be a recommendation of a decent <br>enough hosting service where someone has had success installing <br>Bio-Perl.  I'd try to set up my Mac web sharing feature and then try <br>to setup the stuff locally, but I haven't yet been able to <br>successfully get the port forwarding feature working properly on the <br>apple airport extreme--perplexing.  Next, I might just try to install <br>via the Build.pl script. <br><br>Hmm, checking the wiki, it seems I'll still be able to run remote <br>blast and use the basic seq modules, although some discrepancies and <br>idiosyncrasies may be expected?  Any head-ups about any false <br>assumptions by me would be greatly appreciated. <br><br>Thanks in advance, <br>Roy <br><br>On Fri, Nov 20, 2009 at 5:00 AM, Chris Fields <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453764&i=2" target="_top" rel="nofollow">cjfields@...</a>> wrote: <div class='shrinkable-quote'><br>> <br>> On Nov 20, 2009, at 4:44 AM, Charles Plessy wrote: <br>> <br>>> Le Fri, Nov 20, 2009 at 02:21:54AM -0800, Chu, Roy a écrit : <br>>>> <br>>>> Does anyone use dreamhost as a web hosting service? I'm just curious <br>>>> if anyone has had any luck installing the module as their daemon seems <br>>>> to kill my process whenever I try to install it. Dreamhost tech <br>>>> support attributes it to either exceeding the allocated memory cache <br>>>> or exceeding the processing time. I tried to nice the process, but <br>>>> that didn't help for me. Any luck or experience in resolving this <br>>>> would be much appreciated. I suppose my next attempt would be to try <br>>>> installing it directly and hope I don't need root... <br>>> <br>>> Dear Roy, <br>>> <br>>> DreamHost uses Debian, so you can suggest them to install the Debian package. <br>>> If you are in contact with the tech service, do not hesitate to tell them to <br>>> contact me if they are interested by a backport of the 1.6.0 package. For <br>>> version 1.6.1, it may be more difficult as it depends on perl 5.10.1. <br>> <br>> Any reason why this is so? We specify compatibility back to 5.6.1. <br>> <br>> Alex mentioned the reliance on the specific Extutils::Manifest version. The <br>> version requested has an important bug fix, is present on CPAN, and is <br>> backwards-compatible to 5.6.1. It should be fairly easy to request that as a <br>> separate package. <br>> <br>> A strict requirement for perl 5.10.1 doesn't make sense in that light, unless <br>> said perl maintainer can enlighten us as to why this is an issue? This one may <br>> require a ranty blog post. <br>> <br>>> PS: if you propse to install BioPerl as a feature in the Dreamhost panel, I <br>>> will vote for it :) <br>>> <br>>> Have a nice day, <br>>> <br>>> -- <br>>> Charles Plessy <br>>> Debian Med packaging team, <br>>> <a href="http://www.debian.org/devel/debian-med" target="_top" rel="nofollow">http://www.debian.org/devel/debian-med</a><br>>> Tsurumi, Kanagawa, Japan <br>> <br>> chris <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453764&i=3" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>> </div><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453764&i=4" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453764&i=5" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26453706 ohlohers 2009-11-20T20:17:12Z 2009-11-20T20:17:12Z Mark A. Jensen You can now add your Ohloh widgets and increase your carbon footprint with the less crufty: <br><br> <winterPreWiki> <br> {{#ohloh|acct_id|TYPE}} <br> </winterPreWiki> <br><br>where TYPE is [Detailed|Rank|Tiny]. Taint checks aplenty. <br>MAJ <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453706&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26453147 Excessive CPU use by various Bioperl sites 2009-11-20T17:40:14Z 2009-11-20T17:40:14Z RobertBradbury I run a Linux system which is in a gradual process of evolution from the <br>default Linux browsers (Galeon, Epiphany, etc.) through Firefox (better) to <br>Google's Chromium (IMO, perhaps the best so far).  Chromium allows one to <br>create a process per tab/URL so one can effectively track what it is doing. <br> It also allows one to track the machine usage of these processes (through <br>the Developer > Task manager [shift-escape keyboard] option) which though <br>expensive in terms of overhead allows one to track offending windows (in <br>terms of memory or CPU use).  My processor recently jumped from a typical <br>700 MHz to 1.4 GHz speed (using the Linux Ondemand scheduler - which saves <br>~20 W at the wall outlet -- I've measured it) to the full tilt 2.8 GHz the <br>CPU is capable of.  Looking at the chrome task manager I was not surprised <br>to find the NY Times high on the list (they are pushing content, esp. using <br>Javascript) but much to my dismay the Jalview and Howto:Trees:Bioperl <br>appeared to be high on the list.  Now I am forced to ask myself *why* sites <br>which are simply distributing static information are eating up CPU on my <br>machine!  This is a fundamental flaw in the architecture of the sites -- <br>wherein there should be conscious efforts to minimize user-CPU use (or avoid <br>Javascript entirely).  This would not be a problem if I were using Firefox <br>as I can easily use NoScript to block Javacscript from non-approved sites. <br> But it raises the question of when one should allow Javascript to run (one <br>would "normally" approve academic sites by default) when even the academic <br>sites are abusing my CPU.  There needs to be much greater awareness both on <br>the part of software distributors and software consumers that it is *MY* CPU <br>and *MY* Electricty and *MY* contribution to global warming.  And the <br>developers/distributors should not be sucking down those resources without <br>first saying "May I?" and I have the option of saying "No you may not." <br> There is enough we can do productively (running low homology blast <br>searches) without engaging in endless wheel spinning of Javascripts or <br>looped GIFs. <br><br>Robert <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26453147&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26452770 Re: Installing Bio-perl on dreamhost via CPAN 2009-11-20T17:07:23Z 2009-11-20T17:07:23Z Charles Plessy-9 Le Fri, Nov 20, 2009 at 07:00:45AM -0600, Chris Fields a écrit : <br>> On Nov 20, 2009, at 4:44 AM, Charles Plessy wrote: <br>> > <br>> > DreamHost uses Debian, so you can suggest them to install the Debian <br>> > package.  If you are in contact with the tech service, do not hesitate to <br>> > tell them to contact me if they are interested by a backport of the 1.6.0 <br>> > package. For version 1.6.1, it may be more difficult as it depends on perl <br>> > 5.10.1. <br>> <br>> Any reason why this is so?  We specify compatibility back to 5.6.1. <br><br>Dear Chris, <br><br>you make a good point: although for building we need to either depend on perl <br>5.10.1 or package separately Extutils::Manifest, the resulting bioperl package <br>does not depend on such a high version. Therefore, there is no need for a <br>backport, and the latest Debian package can be installed on Debian stable <br>(5.0/Lenny) system. I just checked the Dreamhost machine on which I happen to <br>have an acces, ‘waratahs’, and it seems to be older, but nevertheless it may be <br>worth asking the admins anyway (with the big drawback that they would have to <br>be asked for each update). <br><br>Have a nice week-end, <br><br>-- <br>Charles Plessy <br>Debian Med packaging team, <br><a href="http://www.debian.org/devel/debian-med" target="_top" rel="nofollow">http://www.debian.org/devel/debian-med</a><br>Tsurumi, Kanagawa, Japan <br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26452770&i=0" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p> tag:old.nabble.com,2006:post-26450030 Re: Installing Bio-perl on dreamhost via CPAN 2009-11-20T12:40:24Z 2009-11-20T12:40:24Z Chris Fields-5 BioPerl is pure perl.  If you believe all dependencies are installed, just unpack the dist to a specific directory and point PERL5LIB at it (for bash): <br><br>export PERL5LIB=/home/USER/bioperl/bioperl-live <br><br>Note that if you plan on doing the same for other bioperl-related modules (ex: bioperl-db) you'll need to add 'lib' to it, as they use a generic Module::Build now. <br><br>export PERL5LIB=/home/USER/bioperl/bioperl-db/lib <br><br>You can also add a 'use lib' directive in your scripts as well.  More at the following link: <br><br><a href="http://www.bioperl.org/wiki/Installing_Bioperl_for_Unix#USING_MODULES_NOT_INSTALLED_IN_THE_STANDARD_LOCATION" target="_top" rel="nofollow">http://www.bioperl.org/wiki/Installing_Bioperl_for_Unix#USING_MODULES_NOT_INSTALLED_IN_THE_STANDARD_LOCATION</a><br><br>chris <br><br>On Nov 20, 2009, at 2:23 PM, Chu, Roy wrote: <br><div class='shrinkable-quote'><br>> "sounds very much like you process was killed for prolonged execution <br>> time, or memory usage. We have a daemon in place that monitors for <br>> processes that take up too much of a shared web server's resources, and <br>> this may have kicked in (and often does when trying to install packages <br>> on a shared server)." <br>> <br>> This was the explanation they had.  Regarding asking their admins to <br>> install, it seems is a "they'll try to get to it but don't hold your <br>> breath situation." <br>> <br>> Hmmm, I tried some other attempts, installing 1.4.0 posed no problems. <br>> I'm not a perl guru, so I tried to increase the build cache size from <br>> the default, 10 MB, hoping that that may be the problem--can't imagine <br>> how though, since I can't imagine how big the whole package version <br>> can differ by (though honestly, I haven't checked). <br>> Whenever I try to install 1.6.1, it runs into a problem I guess after <br>> the 'make' step and lists the <br>> modules--BioPerl-1.6.0/t/Variation/SeqDiff.t <br>> BioPerl-1.6.0/t/Variation/SNP.t <br>> BioPerl-1.6.0/t/Variation/Variation_IO.t <br>> --and typically gets killed here '> Killed' <br>> <br>> Next, I tried 1.6.0, then I get this: <br>> "(I think you ran Build.PL directly, so will use CPAN to install <br>> prerequisites on demand) <br>> CPAN: Storable loaded ok (v2.12) <br>> Going to read '/home/$username/.cpan/Metadata' <br>> Killed" (everything prior works and it seems to get further along than <br>> when I try to install 1.6.1) <br>> <br>> Any insight into why this may be happening would be appreciated. <br>> Something EQUALLY appreciated would be a recommendation of a decent <br>> enough hosting service where someone has had success installing <br>> Bio-Perl.  I'd try to set up my Mac web sharing feature and then try <br>> to setup the stuff locally, but I haven't yet been able to <br>> successfully get the port forwarding feature working properly on the <br>> apple airport extreme--perplexing.  Next, I might just try to install <br>> via the Build.pl script. <br>> <br>> Hmm, checking the wiki, it seems I'll still be able to run remote <br>> blast and use the basic seq modules, although some discrepancies and <br>> idiosyncrasies may be expected?  Any head-ups about any false <br>> assumptions by me would be greatly appreciated. <br>> <br>> Thanks in advance, <br>> Roy <br>> <br>> On Fri, Nov 20, 2009 at 5:00 AM, Chris Fields <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26450030&i=0" target="_top" rel="nofollow">cjfields@...</a>> wrote: <br>>> <br>>> On Nov 20, 2009, at 4:44 AM, Charles Plessy wrote: <br>>> <br>>>> Le Fri, Nov 20, 2009 at 02:21:54AM -0800, Chu, Roy a écrit : <br>>>>> <br>>>>> Does anyone use dreamhost as a web hosting service?  I'm just curious <br>>>>> if anyone has had any luck installing the module as their daemon seems <br>>>>> to kill my process whenever I try to install it.  Dreamhost tech <br>>>>> support attributes it to either exceeding the allocated memory cache <br>>>>> or exceeding the processing time.  I tried to nice the process, but <br>>>>> that didn't help for me.  Any luck or experience in resolving this <br>>>>> would be much appreciated.  I suppose my next attempt would be to try <br>>>>> installing it directly and hope I don't need root... <br>>>> <br>>>> Dear Roy, <br>>>> <br>>>> DreamHost uses Debian, so you can suggest them to install the Debian package. <br>>>> If you are in contact with the tech service, do not hesitate to tell them to <br>>>> contact me if they are interested by a backport of the 1.6.0 package. For <br>>>> version 1.6.1, it may be more difficult as it depends on perl 5.10.1. <br>>> <br>>> Any reason why this is so?  We specify compatibility back to 5.6.1. <br>>> <br>>> Alex mentioned the reliance on the specific Extutils::Manifest version.  The version requested has an important bug fix, is present on CPAN, and is backwards-compatible to 5.6.1.  It should be fairly easy to request that as a separate package. <br>>> <br>>> A strict requirement for perl 5.10.1 doesn't make sense in that light, unless said perl maintainer can enlighten us as to why this is an issue?  This one may require a ranty blog post. <br>>> <br>>>> PS: if you propse to install BioPerl as a feature in the Dreamhost panel, I <br>>>> will vote for it :) <br>>>> <br>>>> Have a nice day, <br>>>> <br>>>> -- <br>>>> Charles Plessy <br>>>> Debian Med packaging team, <br>>>> <a href="http://www.debian.org/devel/debian-med" target="_top" rel="nofollow">http://www.debian.org/devel/debian-med</a><br>>>> Tsurumi, Kanagawa, Japan <br>>> <br>>> chris <br>>> _______________________________________________ <br>>> Bioperl-l mailing list <br>>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26450030&i=1" target="_top" rel="nofollow">Bioperl-l@...</a> <br>>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br>>> <br>> <br>> _______________________________________________ <br>> Bioperl-l mailing list <br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26450030&i=2" target="_top" rel="nofollow">Bioperl-l@...</a> <br>> <a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a></div><br><br>_______________________________________________ <br>Bioperl-l mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26450030&i=3" target="_top" rel="nofollow">Bioperl-l@...</a> <br><a href="http://lists.open-bio.org/mailman/listinfo/bioperl-l" target="_top" rel="nofollow">http://lists.open-bio.org/mailman/listinfo/bioperl-l</a><br><p>From forum: <a href="http://old.nabble.com/Perl---Bioperl-L-f13598.html" embed="fixTarget[13598]" target="_top" >Perl - Bioperl-L</a></p>