Perform full WGS cohort extract scientific tieout for 35 ACMG59 samples #7106
Conversation
| @@ -261,9 +271,9 @@ def compare_sample_data(e1, e2): | |||
| log_difference(key, e1, e2) | |||
|
|
|||
| # TODO: temporary until we decide what to do with spanning deletions | |||
There was a problem hiding this comment.
is this #TODO comment still applicable?
There was a problem hiding this comment.
yep -- tied to https://github.com/broadinstitute/dsp-spec-ops/issues/143, but I'll add that to the comment as well
| POS=$2 | ||
|
|
||
| GVCF=$(cat legacy_wdl/sample_set_membership_v6.tsv | grep $SAMPLE | cut -f2) | ||
| gsutil cat $GVCF | gunzip | grep -C 10 $POS |
There was a problem hiding this comment.
there is a zgrep command that you can use so you don't have to gunzip.
| if((isIndel && totalAsQualApprox < INDEL_QUAL_THRESHOLD) || (!isIndel && totalAsQualApprox < SNP_QUAL_THRESHOLD)) { | ||
| // logger.info(contig + ":" + currentPosition + ": dropped for low QualApprox of " + totalAsQualApprox); | ||
| if ( printDebugInformation ) { |
There was a problem hiding this comment.
comment about totalAsQualApprox. it is the sum of the QualApprox values for every sample with a variant at this site (but not allele specific). this seems strange. it seems like it should be an average or some other ratio.
There was a problem hiding this comment.
It's just the straight sum. I agree it seems like you would want a single high quality variant or some such... but this is exactly what is calculated in the warp pipeline.
| @@ -297,24 +325,19 @@ private double getQUALapproxFromSampleRecord(GenericRecord sampleRecord) { | |||
|
|
|||
| String s = o.toString(); | |||
|
|
|||
| // TODO: KCIBUL -- unclear how QUALapproxes are summed from non-ref alleles... replicating what I saw but need to confirm with Laura | |||
| // Non-AS QualApprox (used for qualapprox filter) is simply the sum of the AS values (see GnarlyGenotyper) | |||
There was a problem hiding this comment.
replace "is simply" but "is approximated by"
…es (#7106) * tieout changes * tidy up, updated README.md * PR comments
Resolves https://github.com/broadinstitute/dsp-spec-ops/issues/239
See README.md in this PR for full details
To make this easier to review, the changes break down into a few sections
Docs -- the README.md. Does it make sense? Could you follow it?
Comparison Script (compare_data.py)-- is it clear? Obvs any bugs would be great. The Github Issue for this PR describes what is compared
WDL changes -- should be straightforward to review, just minor changes
Code changes (java) -- we can walk through this together if that's more effective