[Primeusers] problem running primeGSR

[arve]

Hi,
A common problem is that there are short edges in the species tree. The species tree is discretized, essentially divided up in chunks, and if there are too few chunks on a species edge, is is impossible to place some necessary gene nodes there. 

How many genes and species do you have in your dataset? The present version is unfortunately not tuned for large datasets. Joel Sjöstrand is however working on that. Have you tried a reduced version of your dataset?

	Lars


Mar 25, 2010 kl. 10:16 PM skrev James Cotton:

> We've having difficulty getting PrimeGSR to run - it looks like the  
> input files are read in sucessfully, but before it reports any results  
> from the MCMC, we get a message
> 'Probability: Division with zero attempted!'
> 
> We don't get any output file.
> 
> My initial thought was that this was something to do with the  
> likelihood of some position being effectively zero, but we get this  
> whatever starting parameter values we try, and on input files with  
> different (ultrametric) branch lengths.
> 
> Following suggestions on this forum, I've tried setting -c to a big number (up to 500), and I've tried scaling the tree so the root is at 1.0,  so I'm not sure what else to  try. A command  line is:
> 
> 
> ../primeGSR -o testjac -Sm JC69 -i 100000 -t 100 -Bp 0.001 0.001 -c 150 -Ed Gamma -Hi lyz-species-bl.tre.jac.len1 lyz_code_100323.fa.jac lyz-gene-species-map.gs.jac
> 
> 
> 
> Anything you can do to help would be much appreciated.
> Yours,
> James Cotton
> 
> _______________________________________________
> Primeusers mailing list
> Primeusers at sbc.su.se
> https://mail.sbc.su.se/mailman/listinfo/primeusers