Comparison of eukaryotic phylogenetic profiling approaches using species tree aware methods

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Phylogenetic profiling infers gene function by analyzing gene presence-absence patterns.
• Classical methods show low accuracy in eukaryotes due to genomic complexity.
• Alternative approaches use homologous gene group size correlations.

Purpose of the Study:

• Compare classical presence-absence and group size based phylogenetic profiling in eukaryotes.
• Evaluate performance considering species phylogeny and different gene group categories.
• Assess the accuracy and utility of these methods for inferring functional relationships.

Main Methods:

• Comparative analysis of presence-absence and group size profiles.
• Inclusion of species phylogeny in the comparison.
• Utilized orthology and domain-sharing for gene group categorization.

Main Results:

• Confirmed limited overall performance of phylogenetic profiling in eukaryotes.
• Group size approaches showed initial promise but were overestimated due to simplistic controls.
• Presence-absence profiling proved more accurate for related gene pairs with sufficient information.

Conclusions:

• Presence-absence profiling is a more accurate classifier for gene functional relationships in eukaryotes.
• Group size methods complement presence-absence profiling by detecting co-evolution.
• These methods can estimate functional distances and cluster gene groups, with implications for both eukaryotic and prokaryotic datasets.