Re-Annotation of the Trichomonas vaginalis Genome; Its all about Maverick

Genome annotation projects focus on identifying genes required for host function. Much of the non-genic space consists of low-complexity sequence and transposable elements (TEs). To differentiate host genes from TE genes, repetitive sequence is masked in the assembled genome. The abundance and homogeneity of TEs was one of the difficulties encountered in the annotation of the loblolly pine genome. Another case is the genome of Trichomonas vaginalis. This protozoan human parasite has a 176Mbp genome spread across six chromosomes and is assembled into 17,290 scaffolds. It was reported to have 75,000 protein-coding genes. Studies of the T.vaginalis genome identified a novel TE family, Maverick.  In total Maverick elements account for ~30% of the genome.  Maverick elements are large (15 – 30 kbp), and contain several protein-coding genes that are typically not associated with transposable elements. We hypothesize that Mavericksand other TE encoded proteins have been mis-annotated as genes artificially inflating the gene numbers.

To test this hypothesis we constructed a manually-curated library of the transposable elements from the assembled T.vaginalis genome and used it to mask the published reference assembly. We annotated the masked genome with the genome annotation pipeline, MAKER. This procedure yielded 10,588 genes. Out of the 75,000 protein-coding genes annotated on trichDB.org, almost 27,000 were identified as repeats in these same MAKER outputs. GO term analysis also reveals a significant reduction in TE-related terms (“DNA-binding”, “DNA-integration”, and ”transposase activity”) in the MAKER gene set compared to the original trichDB. set. Collectively these results suggest that the numbers of T.vaginalis protein-coding genes are much more in line with other single-celled eukaryotic parasites than previously believed, and that, in fact, large numbers of repetitive elements, not protein coding genes are the defining characteristic of the T.vaginalis genome.