genotyping step taking a long time

[SaeidehA]

Dear all 

Any suggestion for decreasing delly runtime on genotype calling step? It is taking 36 hours per 30X CRAM file (human data, GRCh38), with 8 CPUs and 8 GB RAM per CPU!

Commands I am using are same as those written in delly GitHub page. 

For genotyping step:

$delly call -g $REF -v $merge_results -o $OUTDIR/$ID.geno.bcf -x $EXC $CRAM

P.S: The reference genome I am using includes decoys and alt contigs. I tried to use a genome without those decoys and contigs, however, delly threw an error for the ref genome, so, I have to use the same reference genome used for aligning the data. 

Thank you in advance,

Best, 

Saeideh

[Tobias Rausch]

Dear Saeideh,

The genotyping performance is mostly dependent on the number of input SVs. I suspect your merge results include several millions of SVs? Can you provide a simple count for each SV type by computing:

bcftools query -f "%INFO/SVTYPE\t%INFO/CT\n" $merge_results | sort | uniq -c

Thanks, Tobias

--
You received this message because you are subscribed to the Google Groups "delly-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to delly-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/delly-users/6a7645d5-aa01-4bfd-8e9e-39ce64b36e9cn%40googlegroups.com.