Small fixes for TRGT #59
Conversation
| if not repeat_id: | ||
| repeat_id = variant_info['info_dict'].get('TRID').split('_')[1] |
There was a problem hiding this comment.
If you like (its only two repetitions so far) a function like
def get_repeat_id(variant_info):
would be welcome, now that its a little bit more code. Besides it lowers apparent complexity with the less C-like and PEPpy.
if this:
do this thing
return this_thing
if that:
do that thing
return that_thing
stranger/utils.py
Outdated
| @@ -236,6 +249,10 @@ def get_trgt_repeat_res(variant_info, repeat_info): | |||
| for allele in mc.split(","): | |||
| mcs = allele.split('_') | |||
| # GT would have the index of the MC in the ALT field list if we wanted to be specific... | |||
|
|
|||
| # What should we do if MC is . ? | |||
There was a problem hiding this comment.
Ah, I hadn't seen that one? Is that from TRGT or sth from STRdust? Do they mean completely missing from motifs, or is it a shorthand for as-reference? ExHu was kind enough to calculate and pass the reference size given the current locus definitions, but I guess for TRGT we would perhaps pass it from the repeat definition file.
We should check what the callers intended here I think!
There was a problem hiding this comment.
This is for TRGT. Probably when the repeat could not be called. Some examples:
chr16 17470908 . GCCGCCGCCGCCGCCC . 0 . TRID=XYLT1;END=17470923;MOTIFS=GCC;STRUC=(GCC)n GT:AL:ALLR:SD:MC:MS:AP:AM .:.:.:.:.:.:.:.
chr16 24613440 . TTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTATTTTAT . 0 . TRID=TNRC6A;END=24613530;MOTIFS=TTTTA,TTTCA;STRUC=(TTTTA)n(TTTCA)n(TTTTA)n GT:AL:ALLR:SD:MC:MS:AP:AM .:.:.:.:.:.:.:.
chr16 66490399 . TAAAATAAAATAAAATAAAATAAAATAAAATAAAATAAAATAAAATAAAATAAAATAATAAAATAAAAA . 0 . TRID=BEAN1;END=66490467;MOTIFS=TGGAA,TAAAA;STRUC=(TGGAA)n(TAAAA)n GT:AL:ALLR:SD:MC:MS:AP:AM .:.:.:.:.:.:.:.
There was a problem hiding this comment.
Ahoy! Yes, could be missing or reality saying hi with some unexpected alleles. I guess we also have to treat it as no-call rather than reference. It's unfortunate that TRGT does seem to set QUAL==0 and no FILTER=='.' for everything. For ExHu we did set the size to 0 for . so I guess we might as well here!
There was a problem hiding this comment.
I run a test-dataset with reads that only covers a few of the pathogenic repeats.
|
I have made some updates the code @dnil, let me know what you think. |
| repeat_id = variant_info['info_dict'].get('REPID') | ||
| if not repeat_id: | ||
| repeat_id = variant_info['info_dict'].get('TRID').split('_')[1] | ||
| repeat_id = get_repeat_id(variant_info) |
|
|
||
| # What should we do if MC is . ? | ||
| if allele == ".": | ||
| repeat_res.extend([0]) |
Co-authored-by: Daniel Nilsson <[email protected]>
|
Let me know if it's ok to merge. Would be great if we could get a bioconda update as well, then I could add Stranger to Nallo (with TRGT for now). |
|
I think this is good to go, but I noticed we have no Changelog-missing warning automation! 😁 |
This PR adds | fixes:
STR_STATUS=no_callor similar?How to prepare for test:
sshto ...bash servers/resources/SERVER.scilifelab.se/update-[THIS_TOOL]-stage.sh [THIS-BRANCH-NAME]How to test:
Expected outcome:
Review:
This version is a: