Matthew's correlation coefficient

I’m interested in comparing clustering methods between average neighbour joining in mothur and a greedy Needleman implemented in CrunchClust, like Schloss et.al. have done recently. But I haven’t been able to figure out how to do that. Any plans for incorporating that into mothur? Or if you did it in R, care to share the code?

Bump.

This would be useful. Not too difficult to implement/calculate from the dist and list files but if some code already exists, it would be nice to not have to write a redundant script…

it’s actually in mothur but takes some serious digging to find

http://www.mothur.org/wiki/Sens.spec

Does this command works in 1.31? I tried a small dataset and it s been runnig for a day without finishing… I tried several settings with no luck either.

I ran this with the 454_SOP files.

mothur > sens.spec(list=final.an.list, column=final.dist, label=0.03)
0.03
It took 25 to run sens.spec.

Output File Names:
final.an.sensspec

Could you post the command you are running?

I probably had a corrupted .dist file. I recalculated it and now it works correctly. Thank you