Parsimony without progress

mothur bugs
1

Hi,

I’m running the parsimony() command on a tree generated by clearcut. I know that the tree file is ok coz it was used to calculate unifrac distances, I also opened it with FigTree and it seems fine. However, when attempting to run the parsimony command nothing happens. the screen shows:
mothur >
parsimony(tree=Biostim.subsample025.phylip.tre, count=Biostim.subsample025.count_table, groups=all, processors=4)

Using 4 processors.
********************###########
Comparing to random:|

and remains this way for hours. CPU usage seems too low to indicate that anything is happening. I also tried to change the no. of processors, limit the groups to two out of the 11 I have, to no avail.

Any ideas?

Thanks,

Daniella

2

The problem is probably that parsimony is trying to calculate the P-value, which requires doing 1000’s of iterations. I think we’ve turned that off by default with the unifrac commands. So you either sit and wait, reduce the number of iterations, or just skip the command.

Pat

3

Hi,

Thanks for the reply, however, sitting and waiting did not pay-up. The program shut down by itself after nearly two days, without a trace of parsimony ever taking place in the LOGFILE. I tried before decreasing iterations to 500, without any result.

Daniella

4

Can you try 2 iterations and see what happens? Also, FWIW, there are better and faster ways of doing this such as AMOVA or HOMOVA.