[maker-devel] CEGMA vs. initial MAKER output for SNAP training

[Carson Holt]

Either works.  Both just count as a starting point (you just need something).  Then you will run MAKER once more with this new HMM and do a final round of bootstrap training using the results of that last MAKER run. Regardless of what you start with, results will converge when you do the last bootstrap training step.

—Carson



> On Aug 25, 2015, at 12:38 PM, John Cornelius <jcornel3 at asu.edu> wrote:
> 
> Hello, I was wondering if it would be better to generate the first hmm for SNAP training from CEGMA results or if I should use the output from my initial MAKER run? Thanks.
> 
> -- 
> John Cornelius
> MCB PhD Candidate
> Arizona State University
> _______________________________________________
> maker-devel mailing list
> maker-devel at box290.bluehost.com
> http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org