search

HKU-BAL / Clair3

Clair3 - Symphonizing pileup and full-alignment for high-performance long-read variant calling
244 stars 26 forks source link

Read Depth in VCF #305

Closed mproberts99 closed 5 months ago

mproberts99 commented 5 months ago

Hi, I found a typo in the output VCF that is causing errors when parsing. The AF parameter has NUMBER=1, so it must have exactly 1 entry, but with the recent update of adding af for all alternate alleles, it has more than one value so it should be NUMBER=G I believe.

FORMAT=

Also, because we use amplicon sequencing, we have to lower the MQ filter to 0 so having the DP automatically display reads with MQ>20 rather than what was specified in the command may be misleading?

zhengzhenxian commented 5 months ago

@mproberts99

Thanks for reporting this! The number of AF has been updated in the header, and we also revised the DP description in the header for clarification. Pls kindly re-pull the latest docker image, thanks!

Zhenxian

mproberts99 commented 5 months ago

Thanks so much!

mproberts99 commented 5 months ago

Ah sorry @zhengzhenxian , it looks like it should actually be NUMBER=R. I got this error: Incorrect number of FORMAT/AF values at chr5:70070208, cannot merge. The tag is defined as Number=G, but found 2 values and 3 alleles. See also http://samtools.github.io/bcftools/howtos/FAQ.html#incorrect-nfields

zhengzhenxian commented 5 months ago

@mproberts99

It seems assigning ‘A’ would be more flexible, I tested with bcftools merge and it worked well. I have updated the docker image and pls kindly re-pull, thanks!