search

bahlolab / cavalier

cavalier: Candidate Variant List Evaluation Report
2 stars 1 forks source link

Test vcf file #1

Closed alig99 closed 7 years ago

alig99 commented 7 years ago

Hi, are you able to supply a test annovar_vcf file? I'm getting this error in R

vars <- load_annovar_vcf(ANNOVAR_vcf, sampleID) Meta line 0 read in. All meta lines processed. Character matrix gt created. Character matrix gt rows: 144 Character matrix gt cols: 8 skip: 0 nrows: 144 row_num: 0

caught segfault address 0x8, cause 'memory not mapped' And I'm assuming it's because my vcf file is incorrect, thanks

mfbennett commented 7 years ago

I have added an example of the script that I use to annotate VCF files using vcfanno and ANNOVAR, which produces annotated VCF file containing all the fields that the load_annovar_vcf function expects to find. I hope this will help illustrate the problem. If not, feel free to email me an example of the VCF file and I would be happy to see where I need to modify this function to make it work in a more general manner.

alig99 commented 7 years ago

Thanks for the example script to annotate VCF files so that load_annovar_vcf function can run.

I then run R & the commands from the cavalier_example.R script. I get as far as getting qualvars. But when I run candvars & then list candvars it says NULL. I did have plenty of variants in qualvars, but then they all get filtered out. All my values in the MAF ESP6500 & MAF 1000G columns are zero. Is it something to do with this? Whereas there are values in MAF ExAC & MAF gnomAD exome & genome columns.

Plus what does the R,A stand for in Proband depth as the first value is either 0 or 1 ?

It would be really useful to have some documentation on how to use Cavalier. I know it's fairly new software, but we are really excited to try to use it as it's capabilities will be very helpful to our work.

mfbennett commented 7 years ago

Edited above comment to remove email signature, etc, and replied via email to answer questions and attempt to fix issues.

Thanks also for great suggestion about wiki.