Next-gen modules

[Chris Fields-5]

On Jul 24, 2009, at 4:28 AM, Peter wrote:

> On Thu, Jul 23, 2009 at 11:58 PM, Chris  
> Fields<cjfields@...> wrote:
>>> i.e. Something like this four line Biopython script would be  
>>> perfect:
>>> http://biopython.org/wiki/Reading_from_unix_pipes
>>
>> We use named parameters so it's a little more verbose.
>>
>> use Bio::SeqIO;
>> my $in  = Bio::SeqIO->new(-fh => \*STDIN, -format => 'fastq-sanger');
>> my $out = Bio::SeqIO->new(-format => 'fastq-solexa');
>> while (my $seq = $in->next_seq) { $out->write_seq($seq) }
>
> Thanks. So that implicitly uses STDOUT for the output?

Yes.

>> Don't be surprised if there are still bugs lurking about, just let  
>> me know
>> and I'll fix 'em.
>
> Have you guys (BioPerl) have also gone for "fastq-sanger" instead of
> just "fastq" for the Sanger Standard version of FASTQ (like EMBOSS)?
> Does BioPerl use just "fastq" to mean anything?

Short answer: yes, and yes.

Slightly longer answer: I've set up SeqIO so it converts "new(-format  
=> 'foo-bar')" to new(-format => 'foo, -variant => 'bar').  In the  
fastq constructor, if the variant is expected but isn't defined (i.e.  
for 'fastq') it defaults to sanger.  Makes it a bit easier maintenance-
wise if a new variant pops up.

> If BioPerl and EMBOSS are using "fastq-sanger", I think Biopython will
> have to support that as an alias too:
> http://lists.open-bio.org/pipermail/biopython-dev/2009-July/ 
> 006416.html
>
> Thanks,
>
> Peter

It's consistent with the 'format-variant' usage, but 'fastq' for us is  
backwards-compatible, so we'll likely support both.

chris

_______________________________________________
Bioperl-l mailing list
Bioperl-l@...
http://lists.open-bio.org/mailman/listinfo/bioperl-l

[Peter-329]

Hi all,

On Fri, Jul 24, 2009 at 1:19 PM, Chris Fields<cjfields@...> wrote:

>>
>> Have you guys (BioPerl) have also gone for "fastq-sanger" instead of
>> just "fastq" for the Sanger Standard version of FASTQ (like EMBOSS)?
>> Does BioPerl use just "fastq" to mean anything?
>
> Short answer: yes, and yes.
>
> Slightly longer answer: I've set up SeqIO so it converts "new(-format =>
> 'foo-bar')" to new(-format => 'foo, -variant => 'bar').  In the fastq
> constructor, if the variant is expected but isn't defined (i.e. for 'fastq')
> it defaults to sanger.  Makes it a bit easier maintenance-wise if a new
> variant pops up.

Right, so BioPerl understands "fastq" and "fastq-sanger" to mean the
Sanger standard FASTQ files.

I've just updated Biopython to also allow "fastq-sanger" as an alias for
"fastq", so we are consistent here:
http://lists.open-bio.org/pipermail/biopython-dev/2009-July/006466.html

Biopython, BioPerl and EMBOSS now all agree on the format names:
* "fastq-sanger" - PHRED scores offset 33
* "fastq-solexa" - Solexa scores offset 64
* "fastq-illumina" - PHRED scores offset 64

And Biopython and BioPerl also agree on the meaning of "fastq" as
an alias for "fastq-sanger". Unfortunately EMBOSS differs here, see:
http://lists.open-bio.org/pipermail/emboss-dev/2009-July/000599.html

Does BioJava or BioRuby have a SeqIO equivalent where they need
to give different sequence formats unique names? If so, we should
talk to them soon...

Peter

_______________________________________________
Bioperl-l mailing list
Bioperl-l@...
http://lists.open-bio.org/mailman/listinfo/bioperl-l

[Joel Martin]

Hello,
   I went to test bioperl-live and Build.PL started updating
modules in my perl install w/o prompting me, frightening!  I
really need to test modules and other groups here need to
test them before they're updated so we don't break anything
when some module's api changes.

   I did svn co of bioperl-live then

perl Build.PL PREFIX=/scratch/bioperl-live

saw some output including
"I think you ran Build.PL directly, so will use CPAN to install prerequisites on demand"

then it went to cpan and tried updating Data::Stag in
my perl install.

How can I ask it to prompt me before updating modules ( so
I can put the updated versions somewhere for it to find
that isn't the live perl install )?  Should I be running
Build.PL indirectly?

Joel
_______________________________________________
Bioperl-l mailing list
Bioperl-l@...
http://lists.open-bio.org/mailman/listinfo/bioperl-l

[Chris Fields-5]

On Jul 24, 2009, at 8:00 AM, Peter wrote:

> Hi all,
>
> On Fri, Jul 24, 2009 at 1:19 PM, Chris Fields<cjfields@...>  
> wrote:
>>>
>>> Have you guys (BioPerl) have also gone for "fastq-sanger" instead of
>>> just "fastq" for the Sanger Standard version of FASTQ (like EMBOSS)?
>>> Does BioPerl use just "fastq" to mean anything?
>>
>> Short answer: yes, and yes.
>>
>> Slightly longer answer: I've set up SeqIO so it converts "new(-
>> format =>
>> 'foo-bar')" to new(-format => 'foo, -variant => 'bar').  In the fastq
>> constructor, if the variant is expected but isn't defined (i.e. for  
>> 'fastq')
>> it defaults to sanger.  Makes it a bit easier maintenance-wise if a  
>> new
>> variant pops up.
>
> Right, so BioPerl understands "fastq" and "fastq-sanger" to mean the
> Sanger standard FASTQ files.

Yes.

> I've just updated Biopython to also allow "fastq-sanger" as an alias  
> for
> "fastq", so we are consistent here:
> http://lists.open-bio.org/pipermail/biopython-dev/2009-July/ 
> 006466.html
>
> Biopython, BioPerl and EMBOSS now all agree on the format names:
> * "fastq-sanger" - PHRED scores offset 33
> * "fastq-solexa" - Solexa scores offset 64
> * "fastq-illumina" - PHRED scores offset 64
>
> And Biopython and BioPerl also agree on the meaning of "fastq" as
> an alias for "fastq-sanger". Unfortunately EMBOSS differs here, see:
> http://lists.open-bio.org/pipermail/emboss-dev/2009-July/000599.html
>
> Does BioJava or BioRuby have a SeqIO equivalent where they need
> to give different sequence formats unique names? If so, we should
> talk to them soon...
>
> Peter

Not sure, but it would be nice to have consistency there, yes.

chris

_______________________________________________
Bioperl-l mailing list
Bioperl-l@...
http://lists.open-bio.org/mailman/listinfo/bioperl-l

[Sendu Bala-2]

Joel Martin wrote:

> Hello,
>    I went to test bioperl-live and Build.PL started updating
> modules in my perl install w/o prompting me, frightening!  I
> really need to test modules and other groups here need to
> test them before they're updated so we don't break anything
> when some module's api changes.
>
>    I did svn co of bioperl-live then
>
> perl Build.PL PREFIX=/scratch/bioperl-live
>
> saw some output including
> "I think you ran Build.PL directly, so will use CPAN to install prerequisites on demand"
>
> then it went to cpan and tried updating Data::Stag in
> my perl install.
>
> How can I ask it to prompt me before updating modules ( so
> I can put the updated versions somewhere for it to find
> that isn't the live perl install )?

Unfortunately, for those few (5) modules that are currently "absolutely
required", it doesn't ask, it just updates.

Data::Stag actually has a comment next to it suggesting it isn't even
"absolutely required". So you could just comment that line out
completely in Build.PL as a temporary fix.

Otherwise, since it will be using CPAN to install modules, you can just
arrange beforehand for CPAN to install to your desired location using
the CPAN configuration system.
_______________________________________________
Bioperl-l mailing list
Bioperl-l@...
http://lists.open-bio.org/mailman/listinfo/bioperl-l

[Chris Fields-5]

On Jul 26, 2009, at 2:33 PM, Sendu Bala wrote:

> Joel Martin wrote:
>> Hello,
>>   I went to test bioperl-live and Build.PL started updating
>> modules in my perl install w/o prompting me, frightening!  I
>> really need to test modules and other groups here need to
>> test them before they're updated so we don't break anything
>> when some module's api changes.
>>   I did svn co of bioperl-live then
>> perl Build.PL PREFIX=/scratch/bioperl-live
>> saw some output including
>> "I think you ran Build.PL directly, so will use CPAN to install  
>> prerequisites on demand"
>> then it went to cpan and tried updating Data::Stag in my perl  
>> install.
>> How can I ask it to prompt me before updating modules ( so
>> I can put the updated versions somewhere for it to find
>> that isn't the live perl install )?
>
> Unfortunately, for those few (5) modules that are currently  
> "absolutely required", it doesn't ask, it just updates.

This is bad; we should never assume the intent of the user.  I would  
rather prompt for these, then bail if the answer is 'no' for any  
'requires' modules.  Leave it up to the user to make the decision to  
update; if they want they will install the latest required modules,  
otherwise there isn't much we can do.

> Data::Stag actually has a comment next to it suggesting it isn't  
> even "absolutely required". So you could just comment that line out  
> completely in Build.PL as a temporary fix.

If you do any work with UniProt, then Data::Stag *is* required. It is  
used by Bio::Annotation::TagTree, the replacement for  
Bio::Annotation::StructuredValue (this has been the case for a couple  
years now I believe).  The Data::Stag update is small but required as  
well, BTW, unless you want warnings popping up.

> Otherwise, since it will be using CPAN to install modules, you can  
> just arrange beforehand for CPAN to install to your desired location  
> using the CPAN configuration system.


chris

_______________________________________________
Bioperl-l mailing list
Bioperl-l@...
http://lists.open-bio.org/mailman/listinfo/bioperl-l