Tonight I continued watching the recording for the ISME19 workshop “From Reads to Function” corresponding to the last part of Day 1. Ben Allen from KBase described completeness and contamination of bins/MAGs. The next step in the workflow is taxonomic assessment and classification. First, one option is to classify taxonomy of metagenomic reads with Kaiju. Reads are assigned by comparison to a reference database containing microbial and viral protein sequences. GOTTCHA2 uses sequence signatures of sequence lineages and includes fungi in its reference database. Insert Genome into Species Tree constructs a species tree using a set of 49 core universal genes defined by the Clusters of Orthologous Groups (COG) gene families. Classify Microbes with GTDB-Tk uses multiple methods to assess and assign taxonomy in comparison to Genome Taxonomy Database reference genomes, including average nucleotide identity scores and placement via topology. The audience asked about dereplication of similar sequences, and Allen noted that it can be used. Another question asked was if there are tools for long reads. The group discussed assemblers, including Meta Spades and which ones work with long reads. An online attendee asked about SQM Tools and assembly, taxonomy, and function with a single tool. It is an R tool, and Elisha noted they would look into it. The group spoke about interoperability and run times needed. The participants also discussed computational power and diversity estimates. I needed this discussion of the different taxonomic assignment tools and their features!
