Estimating PSMC -p parameter and missing heterozygotes

Question

[memory_donk]

Hi Biostars,

My advisor wants me to run PSMC on a data set we've generated (because I'm the only one in the lab who can sort of script and use Linux). I've read the manual on the github page, the original paper (and googled for answers), but I fundamentally don't understand the -p parameter or how to estimate the proportion of missing heterozygotes due to low coverage.

If someone has had to run PSMC any tips would be super helpful. I've already got my fastq consensus and can get psmc to run with the example parameters on the github page (which are probably not suitable for my data). Thanks in advance.

Answer 1

[Zev.Kronenberg]

memory_donk ,

The first thing I would do is try several rates of -p. How much does it affect your Ne estimates?

The second thing you could do it use the lander waterman to take an education guess at your variant calling dropout:

Simulating Read Depth

Last thing to play with is the depth cutoffs.

I've been where you are. Felt your pain.