We are in the final stages of developing a novel algorithm for detecting genes in different species that evolved from the same common ancestor (orthologs). Our algorithm is two orders of magnitude faster than currently used algorithms (OrthoMCL, OrthoFinder, etc.), and has higher overall precision and recall for closely related species.

JustOrthoGroups: Ortholog Grouping Software Package for Large Datasets

We are currently testing a novel ortholog group identification algorithm. Preliminary results suggest that our algorithm is twice as likely as the other algorithms to correctly predict all members of an orthologous group. Furthermore, the clustering algorithm completes tasks that normally take several hours in approximately 1 minute.

We are currently analyzing conserved domains to determine if codon usage biases are conserved across or within domains.

A20 Medicago truncatula genome annotation

We have reconstructed the A20 Medicago truncatula genome, and we are in the process of annotating all of the genes in that species.

We are discovering which codons have the strongest phylogenetic signal when compared against the Open Tree of Life.

We determined a method to recover phylogenetic signal using codon aversion.

We wrote an algorithm for extracting areas of sub-optimal codon usage from a gene sequence.