Yeasts are important for many industrial and biotechnological processes and show remarkable diversity despite morphological similarities. We have sequenced the genomes of 16 ascomycete yeasts of taxonomic and industrial importance including members of Saccharomycotina and Taphrinomycotina. A comparison of these with several other previously published yeast genomes have added increased confidence to the phylogenetic positions of previously poorly placed species including Saitoella complicata, Babjeviella inositovora and Metschnikowia bicuspidata. Phylogenetic analysis also showed that yeasts with alternative nuclear codon usage where CUG encodes serine instead of leucine are monophyletic within the Saccharomycotina. Most of the yeasts have compact genomes with a large fraction of single exon genes with Lipomyces starkeyi and the previously published Pneumocystis jirovecii being notable exceptions. Intron analysis suggests that early diverging species have more introns. We also observed a large number of unclassified lineage specific non-simple repeats in these genomes.

Ascomycete yeasts are metabolically diverse, with great potential for biotechnology. Here, we report the comparative genome analysis of 29 taxonomically and biotechnologically important yeasts, including 16 newly sequenced. We identify a genetic code change, CUG-Ala, in Pachysolen tannophilus in the clade sister to the known CUG-Ser clade. Our well-resolved yeast phylogeny shows that some traits, such as methylotrophy, are restricted to single clades, whereas others, such as l-rhamnose utilization, have patchy phylogenetic distributions. Gene clusters, with variable organization and distribution, encode many pathways of interest. Genomics can predict some biochemical traits precisely, but the genomic basis of others, such as xylose utilization, remains unresolved. Our data also provide insight into early evolution of ascomycetes. We document the loss of H3K9me2/3 heterochromatin, the origin of ascomycete mating-type switching, and panascomycete synteny at the MAT locus. These data and analyses will facilitate the engineering of efficient biosynthetic and degradative pathways and gateways for genomic manipulation.