Unifying the known and unknown microbial coding sequence space

Area of Science:

• Genomics
• Metagenomics
• Molecular Biology

Background:

• A significant portion (40-60%) of predicted genes in microbial systems remain of unknown function, posing a major challenge in molecular biology.
• Existing analytical workflows lack systematic approaches to incorporate and analyze this 'unknown fraction' of genes.

Purpose of the Study:

• To present a conceptual framework and computational workflow (AGNOSTOS) for analyzing genes of unknown function in genomes and metagenomes.
• To quantify the extent, diversity, and relevance of unknown genes across various organisms and environments.
• To demonstrate the utility of analyzing unknown genes for expanding biological understanding and hypothesis generation.

Main Methods:

• Analysis of over 415 million genes from 1749 metagenomes and 28,941 bacterial and archaeal genomes.
• Development and application of the AGNOSTOS computational workflow.
• Compilation of lineage-specific unknown genes from Candidate Phyla Radiation (CPR) bacteria.

Main Results:

• The unknown sequence space is highly diverse, phylogenetically conserved, and taxonomically restricted at the species level.
• A collection of 283,874 lineage-specific unknown genes was identified for CPR bacteria.
• A target gene of unknown function linked to antibiotic resistance was identified, enabling hypothesis generation.

Conclusions:

• The AGNOSTOS workflow effectively bridges the known-unknown gap in genomic and metagenomic data.
• Analysis of unknown genes provides valuable insights into microbial biology, particularly for understudied groups like CPR bacteria.
• This approach can accelerate the discovery of novel functions, including those relevant to practical applications like antibiotic resistance.