Hi,

I am testing bcftools roh on a simulated RAD-seq dataset to see the extent on which bcftools roh can estimate the correct ROH distribution.

I have run the software using the following command:
error_rate=30
recomb_rate=7e-8
input=chrom2.vcf
output=output_ROH_RAD_data.txt

bcftools roh --GTs-only ${error_rate} --estimate-AF - --rec-rate $recomb_rate $input --output $output

The analysis did not conclude after more than 17 hours analysis on 1 node and 20G of RAM memory. I had to stop the running process since I could not figure out if this long running time is expected (due to the sparse distribution of variant sites in RAD-seq datasets or to the low number of SNPs).

What to do you think? What could I do to figure out what is happening?

Here you can find the dataset I have used (downloadable from WeTransfer): https://we.tl/t-92hRTg3Trg
If you do not want to download it, I can post the entire VCF file here.

Thank you in advance.