Orthomcl or CD-HIT for Protein Clustering.

Question

[Naren]

What are basic differences in both the above tools while clustering large datasets of bacterial proteins.

In order to study Pan-genome structure and more specifically unique bacterial genes, which one will be better.

(apart from orthomcl is tedious to install, requires mysql etc.)

[any other tool suggestions are welcome]

[5heikki]

If you want to identify orthologues, in this case reciprocal BLASTPs (or something lighter, e.g. USEARCH) could work too.

Answer 1

[cts]

cd-hit clusters proteins specifically based on the sequence identity. Orthomcl uses a graph-based clustering method to find orthologs without a specific sequence identity cutoffs. Neither one is better, but each is used for a different purpose. cd-hit is used for removing redundancy in a sequence set - like duplicated sequences. Orthomcl is used to find orthologs between species