Genome Annotation of the Cultivated Alfalfa at the Diploid Level (CADL)

Alfalfa (Medicago sativa L.) is an important forage legume species grown in many countries around the world. Since tetraploid alfalfa is widely used for grazing, hay, and silage production, we are interested in annotating a reference genome sequence. With this in mind, we carried out annotation of cultivated alfalfa at the diploid level (CADL). In this way, we hope to optimize our pipeline and tune parameters for alfalfa genome annotation.

The CADL genome assembly, originally sequenced at NCGR with funding from the NSF Medicago truncatula Hapmap Project (IOS-1237993) include 5,753 scaffolds with N50=1.2Mb. The MAKER pipeline was used to identify repeats, align ESTs and proteins to genome, perform ab-initio gene predictions and finally generate gene annotations. A combination of alfalfa data sets including EST sequences, RNA-seq, and protein sequences from related species were used to provide evidence of annotation. The described integrative analysis resulted in 83,077 predicted gene model candidates. The SPADA pipeline was then used to scan the genome for small signal peptide genes. The confidence levels of the predicted gene candidates were estimated using the quality score from MAKER pipeline, expression level from RNA-seq, EST evidences mapping and homologs to its related species Medicago truncatula. 53,972 gene models were further considered as high confidence genes. Further, the quality of our gene model prediction was evaluated using the BUSCO software which discovered 98% of the conservative plant gene orthologs in predicted gene models thus verifying the quality of the assembly and annotation of the CADL genome.