Figure 1

A multi-tiered approach to ePGDB validation. (a) In the absence of highly curated and validated datasets, we took inspiration from the curation-tiered structure of available pathway/genome databases within the BioCyc family. (b/c) Through in silico simulated sequencing experiments on the E. coli K12 genome and two simulated metagenomes, we evaluated the performance of the PathoLogic algorithm under changing sequence coverage and taxonomic distributions. (d) We reanalyzed the genomes of Candidatus Moranella endobia and Candidatus Tremblaya princeps, two symbiotic taxa with reduced genomes, sharing a number of essential amino acid pathways. (e) Finally, we predicted pathways from a previously analyzed paired metagenomic and metatranscriptomic dataset from the Hawaii Ocean Time-series to validate on previously identified pathways and metabolic functions.