Sergei L. Kosakovsky Pond

This paper evaluates the effectiveness of a blinded AI peer review system (BARS) in auditing biomedical literature, finding it improves consistency and detects flaws while highlighting issues with megajournals and template studies.

This paper demonstrates agentic science on public infrastructure to ensure reproducibility by making all agent actions and workflows publicly accessible.

RhizAeon deconvolves reticulate evolution by analyzing sequence mosaics and detecting recombination breakpoints with high accuracy and efficiency.

dFEL extends FEL to model lineage-specific GC/AT drift as per-branch mixtures of three shared regime matrices, correcting selection calls at 1%–4% of boundary codons.

This paper introduces B-STILL, a Bayesian framework to detect functional constraints in genomes by analyzing evolutionary stasis, identifying significant clusters of invariant sites in viral and mammalian genomes.

CAPHEINE automates evolutionary analysis of pathogen sequences, performing selection dynamics and positive selection studies.

This paper shows that accounting for global synonymous selection improves the accuracy of detecting episodic positive selection in protein evolution.

PRIME models physicochemical selection in protein evolution by incorporating biophysical properties, improving model fit and revealing adaptive tuning mechanisms in protein structures.

The paper uses phylogenetic analysis to examine natural selection in zoonotic viruses, finding no evidence of pre-zoonotic adaptation for human transmission but detecting selection changes in the 1977 H1N1 reemergence.

BUSTED-PH identifies genes associated with convergent phenotypes by detecting episodic diversifying selection across mammalian species.

The paper demonstrates a framework using BRC-Analytics and agentic AI to standardize RNA-seq analysis of fungal pathogens, achieving reproducible results across studies with differing annotations.

The paper introduces BRC-Analytics, a browser-based platform for pathogen genomics, demonstrating its use in analyzing measles virus evolutionary dynamics and revealing asymmetric frame evolution under selective pressures.

The paper develops MSS models to estimate selection on synonymous substitutions, finding significant correlations with covariates of codon usage selection across Drosophila species.

The paper identifies that HIV-1 rebound virus primarily originates from a small number of lineages that entered the reservoir near ART initiation, despite the reservoir containing older sequences.

BUSTED-E detects positive selection while identifying residual alignment errors in genome-scale datasets, improving accuracy and biological interpretation.

This study identifies enhanced diversifying selection on polymerase genes in H5N1 clade 2.3.4.4b viruses, linking it to expanded host range and increased pathogenicity.

The paper introduces a heuristic method to adaptively select a genetic distance threshold for HIV transmission network inference, demonstrating its utility in identifying clusters with different risk factors.

The paper produces gapless assemblies of ape sex chromosomes for five great apes and a lesser ape, revealing divergence in Y chromosomes and evolutionary dynamics.

The paper investigates how multicopy gene families on the human Y chromosome survive despite lack of recombination, showing that homogenization via palindromes and arrays, along with purifying selection, preserve protein structure essential for fertility.

The paper identifies proteins and regulatory elements associated with vocal learning by analyzing bat neural activity and epigenomic data.

This study analyzes the evolutionary history of Rem2 across mammals, finding strong purifying selection and epistatic effects that shape its sequence and structure.

The paper develops a framework to account for multi-nucleotide substitutions and synonymous rate variation in natural selection analyses, showing they significantly impact detection of positive selection.

This paper analyzes the evolutionary history of BDNF across 160 mammalian species, finding strict purifying selection in the NGF domain and coevolving sites in the prodomain, linking these findings to disease relevance.

The paper identifies conserved recombination patterns across six coronavirus subgenera and links them to transcriptional regulatory sequences and sequence features.

The paper uses single-cell DNA sequencing to show that defective HIV genomes contribute to phylogenetic inference of reservoir dynamics in an animal model.

TopHap infers phylogenetic structures from common haplotypes in large genome collections, improving accuracy and speed for SARS-CoV-2 variant phylogeny.

RASCL rapidly assesses SARS-CoV-2 clades using molecular sequence analysis to detect selection pressures in emerging variants.

The paper introduces CoV-RDB, a database compiling neutralization data for SARS-CoV-2 variants against monoclonal antibodies, convalescent plasma, and vaccinee plasma, enabling data synthesis and analysis.

The paper analyzes the evolution of the ORF1a/b overlap in coronaviruses, showing it evolved as discrete nucleotide stretches with phylogenetic specificity and is highly conserved across sequences.

The paper presents a global, community-maintained platform for analyzing SARS-CoV-2 genomic data, demonstrating its utility in variant detection and reanalysis of existing datasets.