OrthoHMM: Improved Inference of Ortholog Groups using Hidden Markov Models
Steenwyk, J. L.; Buida, T. J.; Rokas, A.; King, N.
Show abstract
Accurate orthology inference is essential for comparative genomics and phylogenomics. However, orthology inference is challenged by sequence divergence, which is pronounced among anciently diverged organisms. We present OrthoHMM, an algorithm that infers orthologous gene groups using Hidden Markov Models parameterized from substitution matrices, which enables better detection of remote homologs. Benchmarking indicates OrthoHMM outperforms currently available methods; for example, using a curated set of Bilaterian orthogroups, OrthoHMM showed a 10.3 - 138.9% improvement in precision. Rank-based benchmarking using Bilaterian orthogroups and a novel dataset of orthogroups from organisms in three major eukaryotic kingdoms revealed OrthoHMM had the best overall performance (6.7 - 97.8% overall improvement). These findings suggest that Hidden Markov Models improve orthogroup inference.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.