View Single Post
Old 12-19-2014, 05:22 PM   #97
I like code
Location: San Diego, CA, USA

Join Date: Sep 2009
Posts: 436

Yeah they came up with some kind of so-called perfect hash so that every kmer lookup is O(1). The slowest part of their program is probably the EM part (it slows down with increasing size of reference). I think these days if a program intends to put out expression levels for a set of references with ambiguity of mapping between features the EM is the standard. Until something better is figured out...
/* Shawn Driscoll, Gene Expression Laboratory, Pfaff
Salk Institute for Biological Studies, La Jolla, CA, USA */
sdriscoll is offline   Reply With Quote