Develop - #824
Conversation
Filtering, other than transcript lists, or all, can sometimes cause the loss of input genomic variants from output, when G matches a transcript that is later filtered, adjust filtering order a bit to prevent this.
Tweak the shortening of long ref dependant VCF output (deletions, duplications and inversions) to happen on all valid VCF output.
We previously relied on a list of known gapped genes to enable gap mapping correction. This patch switches VV to checking the alignment in the database, and as such needs a database interface upgrade to work, but will negate the issues with maintaining the lists and the overhead of checking non-gapped alignments when only one is affected. Since this would add extra database fetches, which can be the slowest part of the code in the simple case, we also take this as opportunity to switch to only fetching this along with the other from-transcript direction transcript mapping data via a caching store, no more need to repeatedly fetch the exon set.
Vv gapmap insfix alt vcf fix
…invs This commit adds tests and additional functionality to ensure dels, delins, and invs starting in the CDS and spanning into the Ter codon and/or 3 prime UTR apply the hgvs rules that the veriant type is a frameshift
Issue 819
to comply with HGVS rules we have updated our handling of UTR translations to provide p.(=) rather than the old format of p.?
… and 5 prime UTR into CDS
Utr to equality
| ###################################### | ||
|
|
||
| if ( | ||
| ("del" in str(hgvs_naughty.posedit.edit) |
There was a problem hiding this comment.
The logic looks good here, but the same low priority nitpicks apply to this if statement (i.e. hgvs_naughty.posedit.edit.type in ['del', 'delins'] instead of "del" in and == not in for 'ins')
There was a problem hiding this comment.
Some people prefer keeping a clean history for tests, if we want to follow this then merging this commit with the previous commit would be the way to go, I don't mind either way but it can help with some automatic tooling for git bisecting.
John-F-Wagstaff
left a comment
There was a problem hiding this comment.
There is not anything high priority here, so it should be fine to go in as-is, otherwise @Peter-J-Freeman If you want me to have a quick check over after any refinements then that would be fine.
|
Working on the refinements so I know how to use them. Will push up the fixes for a review and any I missed you can polish off |
|
Sounds good, we need to go over the whole code base at some point, but it is such a big job, and should not change the logic, so we haven't had a good enough reason to prioritise it. |
replaces string-based checks with HGVS-native attribute logic. This is a prototype and needs to be rolled out widely
Completes 3 prime UTR handling in MixinInit without stringing
feat: update UTR searching to use datum instead of regex
|
@John-F-Wagstaff should be good to go now. final quick check pls |
|
Looks good, as far as I can see. |
No description provided.