Why the longest insertion is very short

hui-liu commented 4 months ago

I found the DEL size is very long, here are the top ten:

The INV size is also quite long, here are the top ten:

while the INS size is quite short, and here are the top ten:

This is the command I used to call the SVs:

sniffles \
--reference ${ref} \
--input ${bamfile} \
--tandem-repeats ${TR} \
--threads ${th} \
--minsupport 4 \
--minsvlen 50 \
--mapq 20 \
--vcf ${sample}.vcf.gz \
--snf ${sample}.snf \
--sample-id ${sample}

which parameters cause this difference? Do these parameters matter?

--long-ins-length 2500
--long-del-length 50000
--long-inv-length 10000

fritzsedlazeck commented 4 months ago

the insertion sizes are depending on the read length. If an insertion is longer than a read, Sniffles try to combine reads but that not always works out well. In contrast DEL , INV , DUP etc. are not dependent on the read size itself. See our review: https://genomebiology.biomedcentral.com/articles/10.1186/s13059-019-1828-7

hui-liu commented 4 months ago

Thank you!

fritzsedlazeck / Sniffles

Why the longest insertion is very short #500