Fungi in Thailand: A Case Study of the Efficacy of an ITS Barcode for Automatically Identifying Species within the Annulohypoxylon and Hypoxylon Genera

作者：Nuttika Suwannasai, María P. Martín, Cherdchai Phosri, Prakitsin Sihanonth, Anthony J. S. Whalley, John L. Spouge · 发表于：PLoS ONE · 年份：2013 · DOI：10.1371/journal.pone.0054529 · 被引用次数：27 · 研究领域：Plant Pathogens and Fungal Diseases、Mycorrhizal Fungi and Plant Interactions、Yeasts and Rust Fungi Studies

Thailand, a part of the Indo-Burma biodiversity hotspot, has many endemic animals and plants. Some of its fungal species are difficult to recognize and separate, complicating assessments of biodiversity. We assessed species diversity within the fungal genera Annulohypoxylon and Hypoxylon, which produce biologically active and potentially therapeutic compounds, by applying classical taxonomic methods to 552 teleomorphs collected from across Thailand. Using probability of correct identification (PCI), we also assessed the efficacy of automated species identification with a fungal barcode marker, ITS, in the model system of Annulohypoxylon and Hypoxylon. The 552 teleomorphs yielded 137 ITS sequences; in addition, we examined 128 GenBank ITS sequences, to assess biases in evaluating a DNA barcode with GenBank data. The use of multiple sequence alignment in a barcode database like BOLD raises some concerns about non-protein barcode markers like ITS, so we also compared species identification using different alignment methods. Our results suggest the following. (1) Multiple sequence alignment of ITS sequences is competitive with pairwise alignment when identifying species, so BOLD should be able to preserve its present bioinformatics workflow for species identification for ITS, and possibly therefore with at least some other non-protein barcode markers. (2) Automated species identification is insensitive to a specific choice of evolutionary distance, contributing to resolution of a...