Today I found a funny thing, PECAM1/CD31 entry was absent in GenCODE v19.

PECAM1/CD31 is a key surface marker of cell. One of my colleague asked me to check the expression level in one RNA-seq dataset. But I found it is absent in the GenCODE annoation database in the last version based on hg19.

The further investigation indicated that CD31 was in a patched gap (HG183_PATCH, GL383558) of hg19 genome. So it’s not a big surprise for its absence in GenCODE v19. However, it’s ready in the newest hg38 genome and GenCode annotations (currently it’s version 21).

Generally, bioinformaticians would use old genome refence and annotation database 2-3 years after the release of newest genome, for the reseasons of “historical heritage”. But here I need to update the refence genome and annotation database and rerun the RNA-seq pipeline because of the imperfect old annotations.