A recent paper by Forrest et al also provides a catalog of mouse protein kinases as part of a 'phosphatome' (protein kinases + protein phosphatases) analysis . Their catalog lists 561 sequences as protein kinases. We attempted to obtain representative sequences for each of their kinase predictions and compared them to the kinome catalog to determine overlap and reasons for discrepancies.

There are 481 loci in common between the two gene sets. Genes found only in our work inlude most atypical kinases, several divergent kinases, a few of the MARK family expansions and a scattering of other kinases. Sequences found only in the Forrest et al paper include 13 sequences that we classify as psuedogenes, 32 sequences that duplicate previously-counted genes, and 36 additional sequences that do not appear to have clear kinase sequence homology. Some do have matches to the ProSite kinase domain motif (PS00107), which is known to suffer from some false positive predictions. Here are the details: