Figure 1 Pipeline step diagram. Each major component can be fine-tuned with command line options. RAW SEQUENCING FILES (chromatograms) are used for CONSENSUS BUILDING where they are filtered, aligned and converted into a consensus, based on either one of or both forward and reverse sequencing reads. Failure to generate a consensus results in discarding the sample, as the sequence quality is too low. Consensus sequences are then passed to the CLASSIFICATION module where they are BLASTed against the specified database and the BLAST hits are filtered to assign the sample a taxonomic identity. All samples that did not achieve a species-level taxonomic classification undergo a series of QUALITY CHECKS for length, sequence quality and (if the target region includes ITS) the presence of the motifs flanking the ITS region. Samples passing these strict quality checks are sent for CLUSTERING into Operational Taxonomic Units (OTUs). Samples failing the quality checks are directed to SALVAGING where they are BLASTed against a custom database that includes both the reference sequences linked to the samples with species-level assignments and the cluster centroids of the OTUs. All fungal samples that either: (1) have a species-level classification; (2) were clustered into OTUs or (3) were successfully salvaged, can optionally undergo FUNCTIONAL ANNOTATION with FUNGuild (Nguyen et al. 2016).