Will differences in sequencing depth/reads length/insert size affect the results, if all are above 20×?

[kuangzhuoran]

Hi !

I want to compare the demographic histories of different species,
but their sequencing depths vary significantly — from ~20× to ~80×.

What sequencing depth is recommended for running PSMC?

My idea is to downsample all high-coverage data to 20×.

Similarly, the sequencing platforms, read lengths, and insert sizes are also different.

[plubbe]

A general threshold is >18x coverage. I would also apply some filters to remove regions with <5x coverage, and areas more than 2x the mean coverage - doing so will speed up computational time, and removing regions of low coverage will reduce the chance of using incorrectly called bases.

It shouldn't make a big difference if the sequencing platforms, read lengths, inserts are different sizes, especially once they're mapped and filtered and you're moving onto PSMC steps.

Good luck!