The primary differences between the various flavors of the OrthoMCL pipeline are the metrics used to weight the graph and/or MCL inflation parameter value(s).  The effects of these differences can be very difficult to understand and the corresponding publications provide very little, if any, helpful information.  I have therefore systematically explored the effects of these parameters on clustering a subset of sequences from the euKaryotic Orthologous Groups (KOG) database, simulating the fragmentation that commonly results from modern high-throughput sequencing projects.

Upcoming talks: http://umd-byob.github.io/