“…The procedures followed are presented from the scope of analyzing assembled contigs; however, the pipelines also support the annotation of amplicons, fragments, and, in the case of MGnify, unassembled reads, by using most of the same tools. Some specific details differentiate among the workflows, as each may use different tools for the same type of annotation, or perform additional analyses; for example, the DOE JGI pipeline also searches for CRISPR elements ( Anzalone et al, 2020 ; Makarova et al, 2020 ; Nidhi et al, 2021 ; Chavez et al, 2022 ; Katti et al, 2022 ; Wang et al, 2022 ) with CRT-CLI ( Bland et al, 2007 ; Clum et al, 2021 ). However, all three workflows follow, more or less, the same procedure, which consists of the following stages: i) the detection of non-coding RNA (ncRNA) genes, ii) the prediction of protein-coding genes, and iii) functional annotation of proteins and taxonomic assignment.…”