psmc bootstrap plotting?

0

Entering edit mode

6.1 years ago

vulpecula ▴ 30

Hi all,

I have generated 100 bootstraps for PSMC analyses and I am attempting to plot the bootstraps. My initial (non-bootstrap graph) looks as expected, but following several tidbits online, I can't seem to get the bootstrap values to plot properly with the main plot. I also tried plotting a single bootstrap, which also worked and looks nearly identical to the original plot (so I know that the bootstraps were hopefully at least generated properly).

Does anyone have any experience plotting these? Any help is much appreciated :)

Thanks!!

R genome software error • 3.3k views

ADD COMMENT • link updated 3.1 years ago by morgan • 0 • written 6.1 years ago by vulpecula ▴ 30

0

Entering edit mode

Hey I'm having this same problem, did you find a solution in the end?

Cheers, Morgan

ADD REPLY • link 5.4 years ago by momccarthy.3 • 0

0

Entering edit mode

Would anyone mind sharing the code they used to generate the psmc bootstraps? I'd appreciate it, as I'm struggling with that step! Thanks

ADD REPLY • link 3.2 years ago by truebeliever24 ▴ 50

0

Entering edit mode

Hey,

Did you try the commands from (https://github.com/lh3/psmc)? They first use fq2psmcfa to split the original psmcfa file and then run 100 iterations of bootstrapping. The last line is for plotting. Does this work for you?

utils/fq2psmcfa -q20 diploid.fq.gz > diploid.psmcfa

utils/splitfa diploid.psmcfa > split.psmcfa

psmc -N25 -t15 -r5 -p "4+252+4+6" -o diploid.psmc diploid.psmcfa seq 100 | xargs -i echo psmc -N25 -t15 -r5 -b -p "4+252+4+6" -o round-{}.psmc split.fa | sh

cat diploid.psmc round-*.psmc > combined.psmc utils/psmc_plot.pl -pY50000 combined combined.psmc

ADD REPLY • link 3.1 years ago by morgan • 0

Login before adding your answer.