Reimer, L. C. | Search bioRxiv

bioRxiv2024

Predicting bacterial phenotypic traits through improved machine learning using high-quality, curated datasets

Predicting prokaryotic phenotypes - observable traits that govern functionality, adaptability, and interactions - holds significant potential for fields such as biotechnology, environmental sciences, and evolutionary biology. This study leverages machine learning to explore the relationship between prokaryotic genotypes and phenotypes. Taking advantage of the highly standardized datasets in the BacDive database, we modeled eight physiological properties based on protein family inventories, discuss the evaluation metrics, and explore the biological implications of our models. The high confidence values of our predictions highlight the importance of data quality and quantity for a reliable inference of bacterial phenotypes. Our approach yielded nearly 55,000 new data points for approximately 20,000 strains which are published openly in the BacDive database, enriching existing phenotypic datasets and paving the way for future research and analysis. The open-source software generated can readily be applied to other datasets, for example the IMG/M system for metagenomics, as well as different applications, like the assessment of the potential of soil bacteria for bioremediation projects.

bioinformatics↗

bioRxiv2024

Broad diversity of human gut bacteria accessible via a traceable strain deposition system

Numerous bacteria in the human gut microbiome remain unknown and/or have yet to be cultured. While collections of human gut bacteria have been published, few strains have been made publicly available. A major hurdle in making strains publicly available is their deposition to public culture collections. We propose a framework for the bulk-deposition of strains to culture collections, which removes many of the barriers previously identified (www.dsmz.de/bulk-deposit). Using this bulk-deposition system we have created a publicly available collection of human gut isolates. The Human intestinal Bacteria Collection (HiBC) (www.hibc.rwth-aachen.de) contains 340 strains representing 198 species within 29 families and 7 phyla, of which 29 previously unknown species are taxonomically described and named. These included two butyrate-producing species of Faecalibacterium and new dominant species associated with health and inflammatory bowel disease, Ruminococcoides intestinale and Blautia intestinihominis, respectively. Plasmids were prolific within the HiBC isolates, with almost half (46%) of strains containing plasmids, with a maximum of six within a strain. This included a broadly occurring plasmid (pBAC) that exists in three diverse forms across Bacteroidales species. Megaplasmids were identified within two strains, the pMMCAT megaplasmid is globally present within multiple Bacteroidales species. This collection of easily searchable and publicly available gut bacterial isolates will facilitate functional studies of the gut microbiome.

microbiology↗