search

gkno / gkno_launcher

The gkno launcher for executing tools or pipelines
MIT License
32 stars 7 forks source link

Handle tool and pipeline input parsing from JSON where strings can be str or unic #1

Closed chapmanb closed 11 years ago

chapmanb commented 11 years ago

Hi all; I ran into two issues while getting gkno up and running:

Thanks. Looking forward to playing with this more, Brad

chapmanb commented 11 years ago

I also pushed fixes for config_files/pipes/fastq-vcf-baq.json to work with the latest versions of the pipeline, since I was exploring the commandline parameters you used for calmd BAQ. I took the updates from fastq-vcf.json.

AlistairNWard commented 11 years ago

Thanks Brad, this wasn't a problem on the development server that I was using and I just noticed the problem when I tested on some different servers. You saved me the effort of going in and doing the fixes myself though. Thanks for that. I have merged in your changes. Also, since I modified the code, I haven't gone back and updated all the pipelines. The parameters in freebayes have changed and we no longer use BAQ with freebayes, so I need to fix up the pipelines today. Please let me know if there are any other pipelines that you would like building and I'll get them done at the same time.

Al

On Mon, Mar 4, 2013 at 6:36 AM, Brad Chapman notifications@github.comwrote:

I also pushed fixes for config_files/pipes/fastq-vcf-baq.json to work with the latest versions of the pipeline, since I was exploring the commandline parameters you used for calmd BAQ. I took the updates from fastq-vcf.json.

— Reply to this email directly or view it on GitHubhttps://github.com/gkno/gkno_launcher/pull/1#issuecomment-14376527 .

chapmanb commented 11 years ago

Al; Thanks, no worries at all about updating the fastq-vcf-baq pipeline if you don't recommend that approach. At this point what I'd like to do is replicate what you use for human variant calling, so don't need anything special beyond that. Thanks much for merging this all in, Brad

AlistairNWard commented 11 years ago

Ok, I'll talk with Erik and try and get the pipelines updated to reflect the most recent 1000G human variant calling. I'll let you know when it's up.

On Mon, Mar 4, 2013 at 10:39 AM, Brad Chapman notifications@github.comwrote:

Al; Thanks, no worries at all about updating the fastq-vcf-baq pipeline if you don't recommend that approach. At this point what I'd like to do is replicate what you use for human variant calling, so don't need anything special beyond that. Thanks much for merging this all in, Brad

— Reply to this email directly or view it on GitHubhttps://github.com/gkno/gkno_launcher/pull/1#issuecomment-14386640 .

ekg commented 11 years ago

The current method for 1000G calling is just "freebayes -f ref.fa *bam". If I can find any other improvements in default parameters, they will be rolled into the default, so I think this is the best way to do it going forward.

No need for BAQ, possibly no need for BQSR, and no need for indel realignment. Indel realignment is now done internally by default.

On Mon, Mar 4, 2013 at 2:57 PM, AlistairNWard notifications@github.comwrote:

Ok, I'll talk with Erik and try and get the pipelines updated to reflect the most recent 1000G human variant calling. I'll let you know when it's up.

On Mon, Mar 4, 2013 at 10:39 AM, Brad Chapman notifications@github.comwrote:

Al; Thanks, no worries at all about updating the fastq-vcf-baq pipeline if you don't recommend that approach. At this point what I'd like to do is replicate what you use for human variant calling, so don't need anything special beyond that. Thanks much for merging this all in, Brad

— Reply to this email directly or view it on GitHub< https://github.com/gkno/gkno_launcher/pull/1#issuecomment-14386640> .

— Reply to this email directly or view it on GitHubhttps://github.com/gkno/gkno_launcher/pull/1#issuecomment-14401651 .