Inversion OR Duplication ?

[jxlabWzZ]

Hi, I have checked the INV of several super large fragments, and I think it should be a DUP. So how to adjust parameters to judge it as DUP?

[mnshgl0110]

Hi @jxlabWzZ. I agree, this does not seem to be a inversion. Could you please share the syri.out file for the chr10 and the parameters that you used?

[jxlabWzZ]

./miniconda3/envs/syri/bin/syri --version 1.6.3

My script is : ./miniconda3/bin/nucmer --batch 1 -l 50 -c 100 -b 200 -t ${Threads} ref.fa qry.fa ./miniconda3/bin/delta-filter -m -i 90 -l 100 out.delta > ${output}.filtered.delta ./miniconda3/bin/show-coords -THrd ${output}.filtered.delta > ${output}.filtered.coords ./miniconda3/envs/syri/bin/python ./miniconda3/envs/syri/bin/syri -c ${output}.filtered.coords -d ${output}.filtered.delta -r ref.fa -q qry.fa

The coord and syri.out: p20mb.coords.txt p20mb.out.txt

Thanks for your reply!

[mnshgl0110]

Thanks for sharing. Syri seems to be over-fitting these tandem repeat regions. It finds 56 alignments that supports this large inversion, but these alignments have large gaps between them.

$ grep INV1015 p20mb.out.txt | grep INVAL

You can manually filter out such inversions. You can also try setting a smaller value for the --invgaplen option.