Data processing and OTU clustering for ITS amplicons

The ITS targeted ribosomal gene region is longer⁴⁶ than the 16S V4 sequences. An initial analysis indicated that many paired-end reads from ITS libraries did not overlap. This result is a indication that read length might not be enough to cover the entire ITS region of some microbial groups. In order to avoid bias, we decided to use only the first read of the paired-end sequences to perform the ITS amplicon analysis. Single reads were tail trimmed to 175 bp and filtered for a 0.5 expected error threshold using the “fastq_filter” from the USEARCH command. Sugarcane ITS sequences were trimmed using DUK software with parameters “-k 20 -c 2” and a reference sequence obtained from an in-house sugarcane reference assemblage. Reads with no recognizable primer sequences were discarded (primer filter) and the remaining reads were tail primed to the primer sequence (tail trimming). The proceeding steps of dereplicating, discarding singletons and clustering was done using the UPARSE pipeline.