search

fritzsedlazeck / SURVIVOR

Toolset for SV simulation, comparison and filtering
MIT License
354 stars 47 forks source link

SURVIVOR filter using bed #68

Closed Hughes-Shin closed 5 years ago

Hughes-Shin commented 5 years ago

I tried to filter out some SV regions using bed file, but it seemed it didn't work. To check whether it works, I used bed files covering large part of hg38 chr1.

[chr1.bed] (no header) chr1 1 5000000

[vcf] chr1 136934 pbsv.INS.17 G GCAAGGGGCTCGGGCTGACCTCTCTCAGTGTGGGAGGGGCCGGTGTGAGGCAAGGGGCTCACGCTGACCTCTGTCCGCGTGGGAGGGCCGGTGTGAGACAAGGGGCTCGGGCTGACCTCTCTCAGCGTGGGAGGGGCTGGTGTGAGGCAAGGGGCTCGGGCTGACCTCTCTCAGCGTGGGAGGGGCCAGTGTGAGGCAAGGGCTCACACTGACCTCTCTCAGCATGGGAGGGGCCGGTGTGAGA . PASS SVTYPE=INS;END=136935;SVLEN=243 GT:AD:DP:SAC 0/1:5,7:12:2,3,4,3

[cmd line] /Path/to/SURVIVOR filter my.vcf test chr1.bed -1 -1 0 -1 test_filt5.vcf

But it didn't filter any SVs, even though I changed some parameters. Were there any points that I had missed? or was there any problem with my bed format? (For information, I tried to filter vcf files generated by sniffles and pbsv2.1.1 and both have failed.)

Another question, Among filtering before/after SURVIVOR merge vcfs, what would you recommend for population study?

fritzsedlazeck commented 5 years ago

the bed format is tab separated?

For larger studies I would filter before and then sort the file (just in case its not sorted). Subsequently do the merging step. Cheers Fritz

Hughes-Shin commented 5 years ago

Ahhhh, you're right, my bad. I forgot to check tabs. Now it works well. Sorry for bothering you with minor things. Thanks.