Long-Read Metagenome-Assembled Genomes Improve Identification of Novel Complete Biosynthetic Gene Clusters in a Complex Microbial Activated Sludge Ecosystem
Roberto Sánchez-Navarro, Matin Nuhamunada, Omkar S. Mohite, Kenneth Wasmund, Mads Albertsen, Lone Gram, Per Halkjær Nielsen, Tilmann Weber, Caitlin M. Singleton
Abstract
Cataloguing secondary metabolite (SM) potential using genome mining of metagenomic data has become the method of choice in bioprospecting for novel compounds. However, accurate biosynthetic gene cluster (BGC) detection requires unfragmented genomic assemblies, which have been technically difficult to obtain from metagenomes until very recently with new long-read technologies. Here, we determined the biosynthetic potential of activated sludge (AS), the microbial community used in resource recovery and wastewater treatment, by mining high-quality metagenome-assembled genomes generated from long-read data. We found over 4,000 BGCs, including BGCs in abundant process-critical bacteria, with no similarity to the BGCs of characterized products. We show how long-read MAGs are required to confidently assemble complete BGCs, and we determined that the AS BGCs from different studies have very little overlap, suggesting that AS is a rich source of biosynthetic potential and new bioactive compounds.