Alexandros Stamatakis

The paper introduces a new method to discretize Gamma rate distributions in phylogenetics, reducing branch length bias compared to the standard DGM.

RAxML-NG 2.0 introduces new features for phylogenetic inference, including automatic model selection and faster branch support metrics, with significant speed improvements.

The paper evaluates 54 tree shape indices on 45 million empirical trees, revealing rooting and size dependencies, and recommends selecting uncorrelated indices for reliable tree shape analysis.

The paper uses phylogenetic methods to determine the root of the Archaea, finding support for a methanogenic ancestor and a derived origin for host-associated lineages.

This study uses DNA metabarcoding to survey insect biodiversity in Crete, revealing many new species and highlighting gaps in reference databases.

The paper quantifies and predicts multiple sequence alignment difficulty using AlDiScore, achieving high R2 scores on unseen datasets.

This study investigates genetic relationships between the Corinthian metropolis and its colony Amvrakia, finding contributions from both Corinthian migrants and local populations.

This study identifies novel Bacilli endophytes from plants with different lifestyles, revealing extensive biosynthetic potential and functional diversity.

The paper introduces bigrig, a simulator for the DEC model of range evolution, validated with rigorous statistical testing and capable of generating data for large trees quickly.

raxtax is a fast, accurate k-mer-based taxonomic classifier that outperforms existing tools in speed while maintaining accuracy on large datasets.

This paper investigates overfitting in ML phylogenetic inference using cross-validation and finds no significant overfitting in most tools.

This paper addresses non-reproducibility in parallel phylogenetic inference by introducing a bit-reproducible tool and a general algorithm for associative parallel reduction operations.

Pythia 2.0 improves upon its predecessor by using more data and a new model to better predict phylogenetic analysis difficulty, with increased accuracy and speed.

This study evaluates how aDNA damage affects phylogenetic placement accuracy using empirical datasets and finds that read length dominates over other factors.

Pandora estimates the stability of dimensionality reduction in genotype data using bootstrapping and k-means clustering to assess uncertainty in population assignments.

The paper introduces EBG, a machine learning tool that predicts Felsenstein Bootstrap Support values for phylogenies, achieving high speed and accuracy compared to existing methods.

This paper introduces early stopping criteria to accelerate maximum likelihood phylogenetic inference without sacrificing accuracy, achieving faster inference times on empirical and simulated datasets.

The paper introduces a new data structure, genealogical forests, which compresses genealogical trees by storing shared subtrees only once, improving computational efficiency for population genetics analyses.

AleRax estimates gene and species trees together using a probabilistic model of gene duplication, transfer, and loss, achieving higher speed and robustness than existing methods.

This study evaluates the realism of sequence evolution simulations by testing how well supervised learning can distinguish simulated from empirical MSAs.

The paper introduces an adaptive tree search heuristic in RAxML-NG that adjusts search thoroughness based on dataset difficulty, achieving significant speedups while maintaining accuracy.

Lagrange-NG improves the efficiency and quality of ancestral range inference under the DEC model by introducing parallelization and better coding standards.

The paper introduces Pythia, a Random Forest Regressor that predicts the difficulty of phylogenetic analyses under the Maximum Likelihood model.

The paper evaluates phylogenetic inference tools on simulated and empirical data, finding poor performance on difficult alignments and highlighting discrepancies between simulated and empirical accuracy.

Asteroid is a new supertree algorithm that improves robustness to missing data and speed compared to existing methods like ASTRAL and ASTRID.

This study shows that adjusting numerical thresholds in phylogenetic inference software can significantly speed up tree calculations without affecting tree quality.

NetRAX infers maximum likelihood phylogenetic networks without incomplete lineage sorting, efficiently handling larger datasets than existing tools.

CellPhy infers single-cell phylogenies from scDNA-seq data using a maximum likelihood framework, showing robustness to errors and superior accuracy and speed compared to existing methods.

SpeciesRax infers species trees from gene family trees, accounting for duplication, transfer, and loss, and is accurate and faster than existing methods.

RAxML Grove provides a database of over 60,000 empirical phylogenetic trees and model parameters for simulating realistic data and analyzing tree shape distributions.