Naming the untouchable – environmental sequences and niche partitioning as taxonomical evidence in fungi

摘要

The software package DADA2 (version 1.14.0) (Callahan et al. 2016) for R (version 3.6.1) (R Core Team, 2019) was used to quality filter the raw reads and infer ASVs after removing primer sequences with cutadapt (version 2.6 with Python version 3.7.5) (Martin 2011) (Method S4). All 96 samples from the chip were pooled for ASV inference and chimera detection, resulting in a total of 4822 ASVs, representing 1,804,814 reads (37.6% of raw reads). For clarity, ASVs generated from IonTorrent data are called itASVs throughout the text. After removing non-fungal itASVs and itASV_1 (the positive control sequence), the dataset consisted of 4461 itASVs and 1,680,044 reads, of which 282 itASVs and 425,349 reads were putatively Archaeorhizomycetes across the sequencing run. The R package vegan (version 2.5–6) (Oksanen et al. 2013) was used to conduct an nMDS ordination of the fungal itASV occurrence matrix across all samples based on per-sample relative abundances using the Bray-Curtis dissimilarity index (Fig. S3). Finally, the ”ecological” dataset was generated by removing all samples from other studies, controls and all itASVs occurring only once across the 36 Jädraås samples. The count-per-sample matrix of the “ecological” dataset covers 1664 itASVs (619,176 reads) (Supplementary datafile 1), and the sequences are published in GenBank (accession numbers {"type":"entrez-nucleotide","attrs":{"text":"MT926458","term_id":"1896789569"}}MT926458-{"type":"entrez-nucleotide","attrs":{"text":"MT928121","term_id":"1896791573"}}MT928121). Of these, 123 itASVs (233,667 reads; 37.7% of Jädraås fungal reads) were putatively Archaeorhizomycetes (Table S4, Supplementary datafile 2).