Understanding and mitigating some limitations of Illumina (C) MiSeq for environmental sequencing of fungi.

Abstract

ITS-amplicon metabarcode studies using the illumina MiSeq sequencing platform are the current standard tool for fungal ecology studies. Here we report on some of the particular challenges experienced while creating and using a ribosomal RNA gene (rDNA) amplicon library for an ecological study. Two significant complications were encountered. First, artificial differences in read abundances among OTUs were observed, apparently resulting from bias at two stages: PCR amplification of genomic DNA with ITS-region Illumina-sequence-adapted-primers, and during Illumina sequencing. These differential read abundances were only partially corrected by a common variance-stabilization method. Second, tag-switching (or the shifting of amplicons to incorrect sample indices) occurred at high levels in positive mock-community controls. An example of a bioinformatic method to estimate the rate of tag switching is shown, some recommendations on the use of positive controls and primer choice are given, and one approach to reducing potential false positives resulting from these technological biases is presented.