I have some very large Alzheimer’s Disease vcf files, one per chromosome, 808 samples (affected and unaffected patients) in each.

We can stratify those samples by their phenotype.  I want to filter the vcf so that,for example, there are only 25 advanced AD samples (columns)  in the GT matrix,  and display that as its own track in igv. Another track would describe 25 control samples.   In a soon-to-be-submitted new package, igvR,  there will be a method like this:

  displayVcfRegion(igv, chrom, start, end, sample.ids, vcfDirectory)

Revisiting the filterVcf document, I was reminded that filters are applied to rows, not to columns.  So I tried to do column filtering directly (and with zero finesse):

GT <- geno(vcf)$GT  # [1] 406 808
x <- colnames(GT)[sample(1:ncol(GT), 5)]
dim(GT[,x])  # [1] 406   5
geno(vcf)$GT <- x

    # Error in all_dims[, 1L] (from test_igvR.R#143) : incorrect number of dimensions

This error makes complete sense.  

Is there a legitimate way to do this?

Thanks.

 - Paul

Hi Paul,

You can filter by sample with the param,

ScanVcfParam(samples=c("samp1", "samp2", "samp45"))

This isn't by the value in the column, just by column name but it might be what you need. Get the column names with 

samples(scanVcfHeader(myfile))

Val