Effects of missing data on topological inference using a Total Evidence approach

Area of Science:

• Evolutionary Biology
• Phylogenetics
• Computational Biology

Background:

• Macroevolutionary studies require phylogenetic trees incorporating both extant and extinct species.
• The Total Evidence approach combines molecular data from living taxa and morphological data from living and fossil taxa.
• Missing data in Total Evidence matrices can compromise the accuracy of inferred phylogenies.

Purpose of the Study:

• To assess the impact of missing data on phylogenetic tree topologies inferred using the Total Evidence approach.
• To investigate how the proportion of living taxa with no morphological data, missing data in the fossil record, and the number of morphological characters affect phylogenetic inference.
• To determine optimal strategies for data collection to improve phylogenetic accuracy.

Main Methods:

• Simulations were used to evaluate the effects of missing data on tree topologies.
• Phylogenetic trees were inferred from complete and incomplete Total Evidence matrices.
• Topologies from matrices with missing data were compared to those from complete matrices.

Main Results:

• The number of living taxa with morphological data and the total number of morphological characters are more critical than missing fossil data for accurate topology inference.
• Increasing morphological data from extant species enhances the recovery of the "best" tree topology.
• Bayesian inference methods consistently outperformed other methods in tree reconstruction.

Conclusions:

• Sampling effort should prioritize morphological data collection from living species to improve Total Evidence phylogenies.
• Bayesian consensus trees are recommended for fixing tree topology priors in subsequent analyses.
• The study provides insights into optimizing data matrices for robust macroevolutionary inferences.