Fixes for regional BAM re-alignment with bgzipped inputs: glia: out_of_range and tabix errors

[chapmanb]

Eric; This fixes two issues I ran into while using glia for realignment included called variants.

• It avoid erroring out when tabix region retrieval has no variants. The current logic would fail both on tabix retrieval due to a return problem with vcflib (fixed at https://github.com/ekg/vcflib/pull/28) and in regions which don't have any variants where setRegion returns false. The fix checks the logic but avoids erroring out, instead only iterating over variants if they are present.
• Avoids std::out_of_range errors from referenceSequence.at(sp) when an alignment falls over the end of reference sequence. The fix counts these as mismatches.

Here's a small reproducible test case that triggers these errors.

wget https://github.com/chapmanb/bcbio.variation.plus/raw/master/test/data/ensemble/NA12878-10.bam
wget https://raw.github.com/chapmanb/bcbio.variation.plus/master/test/data/ensemble/chr10-start.fa
wget https://s3.amazonaws.com/chapmanb/union-10_250000_400000.vcf.gz
tabix -p vcf union-10_250000_400000.vcf.gz
samtools view -bu NA12878-10.bam | glia -d -Rr -w 1000 -S 200 -Q 200 -G 4 -f chr10-start.fa -v union-10_250000_400000.vcf.gz

Thanks again for this tool. I'm excited to have a minimal test case running with it and looking forward to exploring more.

[ekg]

Hi Brad,

In trying to run the test I hit a snag. The BAM file (https://github.com/chapmanb/bcbio.variation.plus/raw/master/test/data/ensemble/NA12878-10.bam) does not appear to exist, or the URL has changed. Could you please update this and I'll run the tests.

[chapmanb]

Erik; Sorry about this. I moved these over to another repository. The new wgets are:

wget https://github.com/chapmanb/bcbio.variation.recall/raw/master/test/data/ensemble/NA12878-10.bam
wget https://raw.github.com/chapmanb/bcbio.variation.recall/master/test/data/ensemble/chr10-start.fa

Please let me know if you run into anything else. Thanks much for looking at these.

[ekg]

Great, I've just downloaded them.

On Fri, Jan 31, 2014 at 7:22 AM, Brad Chapman [email protected]:

Erik; Sorry about this. I moved these over to another repository. The new wgets are:

wget https://github.com/chapmanb/bcbio.variation.recall/raw/master/test/data/ensemble/NA12878-10.bam wget https://raw.github.com/chapmanb/bcbio.variation.recall/master/test/data/ensemble/chr10-start.fa

Please let me know if you run into anything else. Thanks much for looking at these.

Reply to this email directly or view it on GitHubhttps://github.com/ekg/glia/pull/1#issuecomment-33789855 .

[yifangt]

I was wondering where the "fastq.h" header is or, it may not be needed?

[jollymrt]

Hi Erik,

I am trying to run glia to realign a bam file around indels but getting the following error "Command not found : glia Did you mean ...." how to resolve this issue. I am currently in glia directory and have run make and cmake on both bam and glia. I dont see any executable or dll files after running make. Can you please help.

Regards Jolly

[ekg]

Glia needs to be in a directory in your PATH environment variable. This is a Unix level issue that you can resolve by putting the executable in an expected place or adding the current directory with it to the path.

I should reiterate, use vg rather than glia. It has had hundreds of times the use and developmental effort.

https://github.com/vgteam/vg

On Fri, Aug 10, 2018, 00:07 jollymrt [email protected] wrote:

Hi Erik,

I am trying to run glia to realign a bam file around indels but getting the following error "Command not found : glia Did you mean ...." how to resolve this issue. I am currently in glia directory and have run make and cmake on both bam and glia. I dont see any executable or dll files after running make. Can you please help.

Regards Jolly

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ekg/glia/pull/1#issuecomment-411913589, or mute the thread https://github.com/notifications/unsubscribe-auth/AAI4EYBaeKE2Js46IJUchaartmnwpkvZks5uPLKugaJpZM4BVx3Z .

[jollymrt]

Hi Erik,

Thanks for your response. Basically I need to run a realignment on the bam files that will then be used for our in-house variant caller. I looked up vg but it does not have a separate realignment tool or may be I did not looked up thoroughly. Can you please point if this exists in vg ? Also in glia, I did not see any executable file in the package after running make. Will the file be made in a different folder?

Again thanks for your help.

Regards

On Fri, Aug 10, 2018 at 12:43 AM, Erik Garrison [email protected] wrote:

Glia needs to be in a directory in your PATH environment variable. This is a Unix level issue that you can resolve by putting the executable in an expected place or adding the current directory with it to the path.

I should reiterate, use vg rather than glia. It has had hundreds of times the use and developmental effort.

https://github.com/vgteam/vg

On Fri, Aug 10, 2018, 00:07 jollymrt [email protected] wrote:

Hi Erik,

I am trying to run glia to realign a bam file around indels but getting the following error "Command not found : glia Did you mean ...." how to resolve this issue. I am currently in glia directory and have run make and cmake on both bam and glia. I dont see any executable or dll files after running make. Can you please help.

Regards Jolly

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ekg/glia/pull/1#issuecomment-411913589, or mute the thread <https://github.com/notifications/unsubscribe-auth/ AAI4EYBaeKE2Js46IJUchaartmnwpkvZks5uPLKugaJpZM4BVx3Z> .

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ekg/glia/pull/1#issuecomment-412003374, or mute the thread https://github.com/notifications/unsubscribe-auth/ACaFZjwHlZdqh2CX1ZExNnsxnQhpP5vzks5uPTmRgaJpZM4BVx3Z .

-- Jolly Shrivastava Ph.D (Genetics, Genomics and Bioinformatics) University of California, Riverside

[ekg]

@jollymrt let's continue this conversation here: https://github.com/vgteam/vg/issues/1823