Metatranscriptomics as a tool to identify fungal species and subspecies in mixed communities

Abstract

AbstractHigh-throughput sequencing (HTS) enables the generation of large amounts of genome sequence data at a reasonable cost. Organisms in mixed microbial communities can now be sequenced and identified in a culture-independent way, usually using amplicon sequencing of a DNA barcode. Bulk RNA-seq (metatranscriptomics) has several advantages over DNA-based amplicon sequencing: it is less susceptible to amplification biases, it captures only living organisms, and it enables a larger set of genes to be used for taxonomic identification. Using a defined mock community comprised of 17 fungal isolates, we evaluated whether metatranscriptomics can accurately identify fungal species and subspecies in mixed communities. Overall, 72.9% of the RNA transcripts were classified, from which the vast majority (99.5%) were correctly identified at the species-level. Of the 15 species sequenced, 13 were retrieved and identified correctly. We also detected strain-level variation within theCryptococcusspecies complexes: 99.3% of transcripts assigned toCryptococcuswere classified as one of the four strains used in the mock community. Laboratory contaminants and/or misclassifications were diverse but represented only 0.44% of the transcripts. Hence, these results show that it is possible to obtain accurate species- and strain-level fungal identification from metatranscriptome data as long as taxa identified at low abundance are discarded to avoid false-positives derived from contamination or misclassifications. This study therefore establishes a base-line for the application of metatranscriptomics in clinical mycology and ecological studies.