1) Add unique ID to each structural variant in the merged VCF
2) Generate annotations using AnnotSV before merging with Bedpe
3) Filter sequence features (currently RepeatMasker and Encode's DAC mappability blacklist)
4) Identify splicing deletions and filter variants if they are believed to be from spliced cDNA.
Updated Docker image includes svtools, AnnotSV, and custom scripts for filtering and cDNA contamination.
To-do:
[ ] Incorporate select columns from AnnotSV into bedpe file (following discussion)
[x] Produce 2nd bedpe with non-PASS variants removed
[x] Test with cDNA positive controls and verify accuracy of filtration steps.
[x] Add similar workflow for germline calls
[x] Change TRA to BND so that svtools will handle them appropriately
[x] Add annotation for fusions
[x] Adjust filtering so that calls with a minimum of passing callers are marked as PASS. For example, if a call is made by 3 callers and one of the callers has a filter and the minimum passing callers is 2, the variant should be marked as PASS.
I am thinking to leave AnnotSV as a separate file until we figure out all it's behaviors and how to extract what we want. For now there are some issues such as TRA/BND events not being annotated as expected.
1) Add unique ID to each structural variant in the merged VCF 2) Generate annotations using AnnotSV before merging with Bedpe 3) Filter sequence features (currently RepeatMasker and Encode's DAC mappability blacklist) 4) Identify splicing deletions and filter variants if they are believed to be from spliced cDNA.
Updated Docker image includes svtools, AnnotSV, and custom scripts for filtering and cDNA contamination.
To-do:
TRA
toBND
so that svtools will handle them appropriatelyPASS
. For example, if a call is made by 3 callers and one of the callers has a filter and the minimum passing callers is 2, the variant should be marked asPASS
.