1. Maize PASA assemblies were created through alignment of maize EST and cDNA sequences against AZM and BAC assemblies.
2. Gene predictions were generated based on repeat-masked AZM and BAC assemblies (FGenesh, Augustus, Snap, GlimmerHMM and GeneWise).
3. Alignments of plant proteins and transcript assemblies against AZM and BAC assemblies were created.
4. Gene models were created by combining the above gene evidence from PASA alignments, gene predictions, protein and EST alignments using the combiner program Evidence Modeler (Brian Haas, to be published).
5. Putative functions of gene models were generated based on matches against protein and GO databases.
6. AZM and BAC assemblies, and gene models can be accessed via our maize BLAST server.
7. Gene models and functional annotation can be accessed via our maize gbrowse.

Protein clustering

• Putative paralogous protein family were created based on Pfam domains and novel domains discovered based on sequence similarity during the clustering process.
• Putative paralogous protein families can be accessed here. (It may take several minutes to load this page.)