Comparison of protein coding gene content of yeast and other fungal genomes

Despite the extensive research the exact function many yeast genes remains unknown. Comparisons to other fungal genomes can add power to the genomic analysis by providing the evolutionary context of genes. Our goal is to compare the protein coding gene contents of fungal genomes and relate the differences to the physiological differences between species and taxonomic groups. We have produced consistent Interpro annotations and clustering of protein coding sequences of 16 sequenced fungal genomes of which 8 are yeasts. Our computational system is based on BioPerl scripts and BioSQL schema for storing the sequences and annotations in a relational database. The clustering of protein sequences is done with Tribe-MCL graph clustering software using distances based on Blast E-values. We have discovered that the number of genes belonging to protein clusters having members from all fungal species studied is negatively correlated with genome size, i.e. larger fungal genomes are likely to have more specialized functions not present in species with smaller genomes. In contrast, based on protein clustering Pezizomycotina and Saccharomycotina seem to differ in their level of paralogy, i.e. in number of duplicated genes. In Saccharomycotina average level of paralogy is positively correlated to the size of the genome. In Pezizomycotina, possibly due to Repeat Induced Point mutations (RIP), no clear correlation exists. Using Generic Genome Browser we have created a web-based system that allows the scientists at VTT to easily utilize comparative information in their work. In particular, we link the Interpro entries and clusters so that a user can for a particular protein family browse the neighborhood of the family members detected by Interpro to assess the true extend of the family. In addition, a user can easily find the clusters and Interpro entries that have interesting species distributions. We also link the S. cerevisiae metabolic model iND750 to the comparative data.