Functional association prediction by community profiling

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Advances in sequencing technology have led to a massive increase in DNA and protein sequences, particularly from metagenomics.
• Current methods for metagenome annotation struggle with hypothetical proteins, which often constitute a significant portion of sequences.
• Hypothetical proteins lack known functions, hindering the complete functional profiling of microbial communities.

Purpose of the Study:

• To develop and evaluate a novel community profiling approach for predicting functional associations between proteins.
• To address the challenge of annotating hypothetical proteins in metagenomic data.
• To compare the effectiveness of community profiling against traditional phylogenetic profiling.

Main Methods:

• Proposed a community profiling approach where proteins with similar presence/absence patterns across microbial communities are predicted to be functionally associated.
• Tested various profile construction methods, reference metagenome selections, and correlation metrics to optimize the approach.
• Compared community profiling with phylogenetic profiling using different datasets, including recently sequenced bacterial genomes.

Main Results:

• Community profiling alone slightly outperformed phylogenetic profiling for protein association in well-represented species.
• Combining community and phylogenetic profiling offered marginal improvements in prediction accuracy.
• Community profiling significantly outperformed phylogenetic profiling in revealing more functional associations for a recently sequenced bacterial genome.

Conclusions:

• Community profiling is a valuable method for predicting functional associations and annotating hypothetical proteins in metagenomic data.
• This approach offers advantages over phylogenetic profiling, especially for less-characterized microbial communities.
• Integrating community profiling enhances our ability to understand the functional potential of microbial ecosystems.