cooccurrence with SIM9

Hi Patrick & Sarah, I don’t know if this is a bug, but when I run cooccurrence with default C score metric and matrixmodel 9 the program stalls. This doesn’t happen if I use default setting (model 2) or model 4, which makes me think my .shared input file is probably OK.

thanks,

Giovanni

Hi Giovanni,
Does the program actually crash or does it just seem to hang? Sim9 can take a long time to run on large matrices or if you are doing a lot of iterations.

Thanks,
Kathryn

it hangs; it doesn’t crash. I’ll try with fewer iterations and will post another message if problem persists - thanks.

Kathryn, reducing the number of iters to 1000 made no difference. The only way I found to run program to completion with SIM9 is to use label=. It’s fine to run like that as I don’t need the scores and p values for all labels. thanks.

Giovanni

Could you send your shared and logfile to mothur.bugs@gmail.com so we can try to resolve the bug?

Hi Giovanni,
Thanks for sending your files. I took at look at your shared file and mothur is hanging on label 0.63. This is because there is only 1 OTU at that distance. The sim9 matrix model requires at least 2 OTUs to run properly. We are going to add a warning to mothur and in the future mothur will skip labels that contain only one OTU for this model. Thanks for bringing this to our attention.
Kindly,
Sarah