Informatic Approaches to Annotation

We are using two strategies for predicting functions of unannotated genes:

Prediction of protein structure/function based upon sequence conservation and structures of homologues in Protein Data Bank or consensus fold recognition.

Taking advantage of clustering of genes of related function in prokaryotic genomes to suggest functions of genes of unknown function:

Bacterial Gene of Known Function	Homologue of Arabidopsis Gene of Unknown Function	Bacterial Gene of Known Function	Baterial Gene of Known Function