An improved probability mapping approach to assess genome mosaicism

Area of Science:

• Genomics
• Bioinformatics
• Evolutionary Biology

Background:

• Maximum likelihood and posterior probability mapping are used to visualize the mosaic nature of prokaryotic genomes.
• Posterior probabilities can overestimate support in four-taxon analyses and are sensitive to long branch attraction artifacts.
• Existing methods require improvement in taxon sampling for reliable phylogenetic assessment.

Purpose of the Study:

• To enhance probability mapping for phylogenetic analysis of prokaryotic genomes.
• To improve the reliability of phylogenetic tree support assessment.
• To refine methods for identifying horizontally transferred genes.

Main Methods:

• Extended datasets by including homologs from reference genomes.
• Applied probability mapping using bootstrap support values from these extended datasets.
• Compared results with original maximum likelihood and posterior probability mapping.

Main Results:

• Mapping bootstrap support values from extended datasets yielded results comparable to original methods.
• The use of bootstrap values provided a more conservative assessment of phylogenetic support.
• This approach facilitates focusing on protein families with strong phylogenetic disagreement.

Conclusions:

• Posterior probability is a non-conservative measure; its mapping offers a preliminary estimation of phylogenetic content.
• Enhanced taxon sampling and subtree analyses mitigate inconsistencies of four-taxon analyses while retaining visualization power.
• Individual multi-taxon phylogeny inspection is crucial for distinguishing horizontal gene transfer from other evolutionary events.