merging CD-HIT results

[Tectona]

Hi:

I have three datasets of sequences that I have clustered with CD-HIT. I would like to do a clustering of the three datasets combined, but that task is too big for my computer. There are instructions in the CD-HIT manual for compare /adding new seuences to an already clustered database (CD-HIT-2D), but that way my computer also runs out of memory.

I wonder if there is a way to compare / merge the three fasta and cluster results from CD-HIT into one.. or whether there is a software tool that is less memory intensive, so I can combine the three original sequence files and cluster them all at once.

Thanks in advance for any suggestions

[Torst]

I think you could use the "clstr_merge.pl" script that comes with cd-hit?

http://weizhong-lab.ucsd.edu/cd-hit/wiki/doku.php?id=cd-hit_user_guide#incremental_clustering

[Tectona]

The clstr_merge.pl script probably can merge the cluster results, as given in the example for the "incremental clustering"... but what about comparison and merging of the fasta results?