Anne‐Florence Bitbol

The paper shows that hydrodynamic flow in the gut reduces the fixation probability of beneficial mutants but speeds up their invasion.

This study shows that cooperation is promoted in deme-structured populations under growth-merging dynamics with hard selection, but not under soft selection.

This study identifies how the intestinal microbiota drives the evolution of pathoadaptive traits in ESBL E. coli through adhesion and biofilm formation.

The paper introduces ProteomeLM-HPI, a language model adapted to predict cross-species host-pathogen interactions using proteome context and fine-tuning.

ProteomeLM predicts protein-protein interactions and gene essentiality across species using a proteome-scale language model, outperforming existing methods in speed and accuracy.

The paper shows environment heterogeneity can amplify natural selection and speed up mutant fixation in graph-structured populations, depending on migration rates and mutant fitness advantages.

The paper shows that fluctuating selection pressures improve coevolution-based inference of protein structure and function from sequence data.

The paper characterizes over 250 Rubisco variants from bacteria and archaea, showing they rarely exceed 30 reactions per second at 30°C and predicting rates for all sequenced variants using machine learning.

RAG-ESM improves pretrained protein language models by incorporating homology information through retrieval-augmentation, achieving better performance in sequence prediction tasks.

ProtMamba is a homology-aware, alignment-free protein language model that outperforms state-of-the-art models in homolog-conditioned sequence generation.

The paper shows spatial structure in bacterial populations increases the probability of evolutionary rescue by antibiotic resistance through enhanced fixation of resistant mutants in small subpopulations.

The paper examines how spatial population structure and migration affect adaptation in rugged fitness landscapes, finding that suppression of selection can enhance early adaptation but reduce long-term fitness.

DiffPaSS improves pairing of interacting protein sequences using differentiable optimization, outperforming existing methods on benchmark datasets.

The paper evaluates how phylogeny affects the identification of functional sectors in protein sequences and shows that two methods, conservation and ICOD, are robust to phylogeny and provide complementary insights.

The paper bridges evolutionary graph theory and experimental conditions by introducing a new model of spatially structured populations and analyzing recent experiments to test predictions about mutant fixation.

This paper introduces DiffPALM, a method to pair interacting protein sequences using masked language modeling, improving structure prediction of protein complexes.

This study systematically analyzes CDR contacts and sequence constraints between TCR αβ chains using TCR structures and mutual information, showing sequence dependencies and co-evolution patterns.

The paper combines phylogeny and coevolution to improve the inference of interaction partners among paralogous proteins, showing significant performance gains in challenging scenarios.

The paper demonstrates that MSA Transformer can generate protein sequences that perform as well as natural sequences in homology, coevolution, and structure-based measures, outperforming Potts models in some cases.

The paper investigates how phylogenetic correlations affect structural contact inference from protein sequences and shows global methods like Potts models are more resilient to these effects.

This paper shows that protein language models using MSA column attentions capture phylogenetic relationships and distinguish coevolutionary signals from historical contingency.

The paper shows that structural and phylogenetic correlations combine to improve protein partner inference from sequences, with non-contact site pairs contributing significantly.

This paper introduces a method to extract the phylogenetic dimension of coevolution, revealing hidden functional signals in proteins and improving structural-contact prediction.

The paper introduces YeaZ, a CNN for accurate yeast cell segmentation without fluorescent markers, outperforming existing methods and revealing new biology in cell morphology.

The paper evaluates how dataset size and quality affect protein-protein interaction inference using synthetic data and shows quadratic scaling relationships.

The paper shows that phylogenetic correlations alone can be used by DCA-based methods to infer protein interaction partners, performing as well as sequence similarity methods in some cases.

The paper introduces a method to identify protein interaction partners from sequence data by maximizing mutual information, showing slight improvement over previous methods.