Table 1 Databases used to build OL2
ProgramVersion in OL1 (date)Version(s) in OL2 (date)No. of C. elegans genes in OL1No. of C. elegans genes in OL2 (% change)No. of human ENSG IDs in OL1No. of human ENSG IDs in OL2 (% change)a
Ensembl Compara57 (2010)87–89 (2016–2017)64046801 (+6.2%)86429186 (+6.3%)
HomoloGene64 (2009)68 (2014)41273778 (−8.5%)29563205 (+8.4%)
InParanoid7 (2009)8 (2013)55915581 (−0.2%)75278949 (+18.9%)
OrthoMCL4 (2010)5 (2011)56635699 (+0.6%)74177588 (+2.3%)
OMANA1 (2016)NA3882NA4558
OrthoInspectorNA2 (2015)NA5361NA7771
  • The programs used here all scored highly in a recent assessment of orthology-prediction methods (Altenhoff et al. 2016). For the four previously used programs, we report the net change (%) in C. elegans and human genes predicted to be orthologs between versions. (For the other two programs, these measurements are not applicable).

  • a The change in human ENSG IDs upon updates includes those whose original IDs were retired, but which still exist in the Ensembl database with a new, unlinked, ID. This deficiency in annotation makes it impossible to assess the true extent of gains and losses in the human gene set (see main text).