Merging vcf files error with glnexus:v1.2.7

[poddarharsh15]

**Have you checked the FAQ?

Describe the issue: Merging vcf files error. Setup

• Operating system: working on cluster
• DeepVariant version:latest
• Installation method (Docker):
• Type of data: (GIAB AshkenazimTrio [HG002,HG003,HG004] analysis.)

Steps to reproduce:

• Command:

udocker run \
-v "${PWD}/output":"/output" \
quay.io/mlin/glnexus:v1.2.7 \
/usr/local/bin/glnexus_cli \
--config DeepVariant_unfiltered \
/output/HG002.g.vcf.gz \
/output/HG003.g.vcf.gz \
/output/HG004.g.vcf.gz \
| udocker run -i google/deepvariant:deeptrio-"${BIN_VERSION}" \
  bcftools view - \
| udocker run -i google/deepvariant:deeptrio-"${BIN_VERSION}" \
  bgzip -c > output/HG002_trio_merged.vcf.gz

• Error trace: (if applicable)

Num BCF records read 118736378 query hits 14552613 [E::bgzf_read_block] Invalid BGZF header at offset 265038798 [E::bgzf_read] Read block operation failed with error 2 after 0 of 32 bytes [E::bgzf_read] Read block operation failed with error 3 after 0 of 32 bytes Error: BCF read err

[akolesnikov]

Looks like this is the GLnexus question. Could you please post the question at GLNexus page Also, from the log output it looks like GLnexus was completed successfully.

[pichuan]

Hi @poddarharsh15

Actually, can you go back in your log and confirm that DeepTrio runs actually finish correctly?

If I remember correctly, our run_deeptrio one-step script might continue to run the following steps even when previous steps failed.

[pichuan]

And, follow up on @akolesnikov 's point, if you have gotten to this point, it would seem like these files should be complete?

/output/HG002.g.vcf.gz \
/output/HG003.g.vcf.gz \
/output/HG004.g.vcf.gz \

If you can examine those files and confirm, that will be great. (Or look at the log like I mentioned before. But given you have the files, checking the files directly might be easier :))

[poddarharsh15]

Hi @pichuan, @akolesnikov,

I'm new to DeepTrio and couldn't locate the log files, but I have intermediate results showing that DeepTrio ran successfully without errors. Additionally, I successfully benchmarked the .vcf files generated by DeepTrio. I've attached screenshots for reference. Your assistance is greatly appreciated. Thank you

finished log

Benchmark

[pichuan]

Hi @poddarharsh15 , it seems like you're certain that the DeepTrio run finished correctly. In that case, I agree with @akolesnikov 's original assessment that this can be an issue for the downstream glnexus step, which we can't directly support.

One suggestion to try: If you need to check your run a bit more closely, maybe breaking it down to just running this part first:

udocker run \
-v "${PWD}/output":"/output" \
quay.io/mlin/glnexus:v1.2.7 \
/usr/local/bin/glnexus_cli \
--config DeepVariant_unfiltered \
/output/HG002.g.vcf.gz \
/output/HG003.g.vcf.gz \
/output/HG004.g.vcf.gz

before piping to the next step. Maybe that could help you identify what the errors are coming out from that step?

[akolesnikov]

Closing the issue. Feel free to reopen as needed.