nQMaker: Estimating Time Nonreversible Amino Acid Substitution Models.

Area of Science:

• Phylogenetics and evolutionary biology
• Computational biology and bioinformatics

Background:

• Amino acid substitution models are crucial for phylogenetic analysis of protein sequences.
• Current models are time-reversible for computational ease, but this is biologically unrealistic and hinders rooted tree inference without outgroups.

Purpose of the Study:

• To introduce nQMaker, a maximum likelihood method for estimating time nonreversible amino acid substitution models.
• To enable the inference of rooted phylogenetic trees using these nonreversible models.
• To improve the accuracy and biological realism of phylogenetic analyses.

Main Methods:

• Developed nQMaker, an extension of the QMaker method, for maximum likelihood estimation of nonreversible amino acid substitution models.
• Applied nQMaker to diverse empirical protein sequence alignments from various taxa (mammals, birds, plants, fungi).
• Integrated nQMaker into the IQ-TREE software for user accessibility.

Main Results:

• Nonreversible models estimated by nQMaker provide a significantly better fit to empirical alignments compared to existing reversible models.
• Model fit improvement scales with the size of the data set.
• nQMaker successfully recovered established root placements for plant and bird trees with high statistical support, eliminating the need for outgroups.

Conclusions:

• nQMaker offers a more biologically realistic and computationally efficient approach to phylogenetic inference.
• The ability to estimate nonreversible models and infer rooted trees without outgroups represents a significant advancement in phylogenetic analysis.
• nQMaker is now available in IQ-TREE, empowering researchers to conduct more accurate protein sequence analyses.