Segmentation fault

[sunny kevin]

Hi,

I'm running codeml analysis for 3500 orthologs.

Using ndata = 10, I'm able to run only for 10 orthologs at a time.

If I increase the ndata option more than 10, codeml ends up in segmentation fault.

I'm performing analysis in my linux server, which had 96 cpus.

Can any one experience similar problem.

Is their any bash or python or perl scripts available online which can automate the codeml analysis in one go.

Please do send the scripts weblinks.

Need your suggestions.

Thanks

Kevin

[Ziheng]

there have been some recent changes to the ndata option.  now the option assumes one tree from the tree file for each alignment in the sequence file.  perhaps you can download a copy and look at the examples/ndata folder, and see whether the new option fixes your problem.

best, ziheng