Fast and accurate bootstrap confidence limits on genome-scale phylogenies using little bootstraps

Area of Science:

• Phylogenetics
• Computational Biology
• Bioinformatics

Background:

• Felsenstein's bootstrap is standard for assessing confidence in species phylogenies from sequence alignments.
• Standard bootstrap involves computationally intensive resampling of alignment sites.
• High computational burden limits its application in large phylogenomic datasets.

Purpose of the Study:

• Introduce a computationally efficient bootstrapping method for phylogenetics.
• Compare the effectiveness of the new method against standard bootstrap.
• Enhance the feasibility of phylogenomic analyses with large datasets.

Main Methods:

• Developed the "bag of little bootstraps" approach for phylogenetics.
• Involves bootstrapping small subsets of sites rather than entire alignments.
• Aggregated results from multiple small bootstrap samples to estimate confidence.

Main Results:

• The "bag of little bootstraps" yields confidence limits comparable to standard bootstrap.
• Achieved similar confidence in species relationships with a fraction of computational time and memory.
• Demonstrated significant improvements in efficiency and reduced computational load.

Conclusions:

• The "bag of little bootstraps" provides a rigorous and efficient alternative to standard bootstrap.
• This method can enhance the rigor, efficiency, and parallelization of big data phylogenomic analyses.
• Facilitates more accessible and scalable phylogenetic inference.