R. Monasson

Deciphering the code of viral-host adaptation through maximum entropy models

This paper introduces a method to decode host and viral family fingerprints in nucleotide usage, showing it outperforms deep learning for classification and enables manipulation of nucleotide composition in viral sequences.

Infer global, predict local: quantity-quality trade-off in protein fitness predictions from sequence data

The paper investigates whether more data or higher-quality data leads to better protein fitness predictions, showing that simple models with carefully selected data can outperform complex models trained on all data.

Generative and interpretable machine learning for aptamer design and analysis of in vitro sequence selection

This study uses Restricted Boltzmann Machines to analyze SELEX data for thrombin aptamers, enabling sequence fitness estimation, classification, and novel molecule generation, which is validated experimentally.

Origins and breadth of pairwise epistasis in an α-helix of β-lactamase TEM-1

The paper quantifies epistasis in a β-lactamase α-helix using double mutants and shows it can be explained by a non-linear two-state model.

Quantitative modeling of the effect of antigen dosage on B-cell affinity distributions in maturating germinal centers

This study models how antigen dosage affects B-cell affinity distributions in germinal centers and shows non-monotonic relationships between antigen levels and average affinity.

Evolution-based design of chorismate mutase enzymes

This paper demonstrates an evolution-based design process for synthetic enzymes using sequence data, showing that statistical models can generate functional sequences with substantial diversity.

Parameters and determinants of responses to selection in antibody libraries

The paper quantifies how scaffold evolutionary maturation affects antibody library response to selection, finding maturated scaffolds show lower response to arbitrary targets.