Fig. 1

Analysis workflow of HumanMycobiomeScan. a Inputs are single- (.fastq) or paired-end (.fastq or compressed .fastq) reads. b Candidate fungal reads are screened by mapping onto reference fungal genomes contained in a precompiled database. This allows for a first reduction of the sample size, lowering the number of sequences that will be subjected to further steps. c Three filtration steps are carried out to eliminate low quality reads as well as reads belonging to humans and bacteria. d The remaining sequences are realigned onto the fungal genome database for definitive taxonomic assignment of the reads. The results are tabulated as both abundance profiles and read counts, and represented by bar plots