arXiv subjects

Aurelien Pelissier

Publications and source records attributed to Aurelien Pelissier.

6 recordsLinked to original sources

arXiv2026

Contextualizing Biological Language Models across Modalities via Logit-Space Contrastive Alignment

Pretrained biological language models expose per-token probability distributions through masked-token prediction, providing the likelihood interface central to sequence design, variant scoring, and mechanistic interpretation. Yet these distributions are learned from broad unlabeled corpora and are not naturally conditioned on task-specific biological contexts such as interaction partners, cellular environments, or therapeutic interventions. Existing contextual matching methods often distort this interface through pooled embeddings, contrastive latent spaces, or task-specific prediction heads. We introduce LOGICA (Logit-space Contrastive Alignment), a framework for context-conditioned prediction that performs contrastive learning directly in output-logit space. Using gated cross-modal adapters compatible with each model's native token head, LOGICA preserves the pretrained likelihood interface and converts contextualized token log-likelihoods into matching scores. Alignment is defined through context-sensitive token probabilities rather than proximity in a shared embedding space, enabling learning from sparse paired data across models with distinct vocabularies, without a shared tokenizer or decoder. LOGICA is particularly effective for mutation-local variant ranking, where comparisons reduce to context-conditioned likelihoods of mutant tokens at perturbed sites. Across protein--ligand binding, TCR--peptide activity, and drug-conditioned resistance prediction, LOGICA improves over prior state-of-the-art methods, including matched latent-contrastive and conditional MLM baselines, while retaining a token-level interface for interpretation and generation. On held-out-gene single-mutation drug-resistance prediction, LOGICA improves AUC from near-random latent-space baselines of $\sim$0.55 to $\sim$0.65.

cs.LG↗

arXiv2026

AI Developments for T and B Cell Receptor Modeling and Therapeutic Design

Artificial intelligence (AI) is accelerating progress in modeling T and B cell receptors by enabling predictive and generative frameworks grounded in sequence data and immune context. This chapter surveys recent advances in the use of protein language models, machine learning, and multimodal integration for immune receptor modeling. We highlight emerging strategies to leverage single-cell and repertoire-scale datasets, and optimize immune receptor candidates for therapeutic design. These developments point toward a new generation of data-efficient, generalizable, and clinically relevant models that better capture the diversity and complexity of adaptive immunity.

q-bio.BM↗

arXiv2025

AbRank: A Benchmark Dataset and Metric-Learning Framework for Antibody-Antigen Affinity Ranking

Accurate prediction of antibody-antigen (Ab-Ag) binding affinity is essential for therapeutic design and vaccine development, yet the performance of current models is limited by noisy experimental labels, heterogeneous assay conditions, and poor generalization across the vast antibody and antigen sequence space. We introduce AbRank, a large-scale benchmark and evaluation framework that reframes affinity prediction as a pairwise ranking problem. AbRank aggregates over 380,000 binding assays from nine heterogeneous sources, spanning diverse antibodies, antigens, and experimental conditions, and introduces standardized data splits that systematically increase distribution shift, from local perturbations such as point mutations to broad generalization across novel antigens and antibodies. To ensure robust supervision, AbRank defines an m-confident ranking framework by filtering out comparisons with marginal affinity differences, focusing training on pairs with at least an m-fold difference in measured binding strength. As a baseline for the benchmark, we introduce WALLE-Affinity, a graph-based approach that integrates protein language model embeddings with structural information to predict pairwise binding preferences. Our benchmarks reveal significant limitations in current methods under realistic generalization settings and demonstrate that ranking-based training improves robustness and transferability. In summary, AbRank offers a robust foundation for machine learning models to generalize across the antibody-antigen space, with direct relevance for scalable, structure-aware antibody therapeutic design.

q-bio.BM↗

arXiv2022

Practical and scalable simulations of non-Markovian stochastic processes and temporal networks with individual node properties

Discrete stochastic processes are prevalent in natural systems, with applications in physics, biochemistry, epidemiology, sociology, and finance. While analytic solutions often cannot be derived, existing simulation frameworks can generate stochastic trajectories compatible with the dynamical laws underlying the random phenomena. Still, most simulation algorithms assume the system dynamics are memoryless (Markovian assumption), under which assumption, future occurrences only depend on the system's present state. This enables efficient and exact simulation via the Gillespie algorithm. However, many real-world systems are inherently non-Markovian and exhibit memory effects. Such systems are difficult to study analytically, and current numerical methods are often computationally expensive or limited by strong simplifying assumptions that conflict with empirical data. To address these limitations, we introduce the Rejection-based Gillespie algorithm for non-Markovian Reactions (REGIR), a general and scalable framework for simulating non-Markovian stochastic systems with arbitrary inter-event time distributions. REGIR provides user-defined accuracy while preserving the same asymptotic computational complexity as the classical Gillespie algorithm. We derive a lower bound on REGIR's approximation accuracy and demonstrate its capabilities across three representative classes of non-Markovian systems: (1) reaction channels with delays, (2) stochastic processes driven by individual reactant properties, and (3) temporal networks governed by node activity. In all cases, REGIR accurately captures memory-dependent dynamics and outperforms existing approaches in terms of flexibility and efficiency.

q-bio.QM↗

arXiv2019

Feature selection as Monte-Carlo Search in Growing Single Rooted Directed Acyclic Graph by Best Leaf Identification

Monte Carlo tree search (MCTS) has received considerable interest due to its spectacular success in the difficult problem of computer Go and also proved beneficial in a range of other domains. A major issue that has received little attention in the MCTS literature is the fact that, in most games, different actions can lead to the same state, that may lead to a high degree of redundancy in tree representation and unnecessary additional computational cost. We extend MCTS to single rooted directed acyclic graph (SR-DAG), and consider the Best Arm Identification (BAI) and the Best Leaf Identification (BLI) problem of an expanding SR-DAG of arbitrary depth. We propose algorithms that are (epsilon, delta)-correct in the fixed confidence setting, and prove an asymptotic upper bounds of sample complexity for our BAI algorithm. As a major application for our BLI algorithm, a novel approach for Feature Selection is proposed by representing the feature set space as a SR-DAG and repeatedly evaluating feature subsets until a candidate for the best leaf is returned, a proof of concept is shown on benchmark data sets.

cs.LG↗

arXiv2019

Beyond the Nucleus: Cytoplasmic Dominance in Follicular Thyroid Carcinoma Detection Using Single-Cell Raman Imaging Across Multiple Devices

Cytological diagnosis of follicular thyroid carcinoma (FTC) is one of major challenges in the field of endocrine oncology due to absence of evident morphological indicators. Morphological abnormalities in the nucleus are typically key indicators in cancer cytopathology and are attributed to a range of biochemical alterations in nuclear components. Consequently, Raman spectroscopy has been widely used to detect cancer in various cytological samples, often identifying biochemical changes prior to observable morphological alterations. However, in the case of FTC, cytoplasmic features such as carotenoids, cytochromes, and lipid droplets have shown greater diagnostic relevance compared to nuclear features. This study leverages single-cell Raman imaging to explore the spatial origin of diagnostic signals in FTC and normal thyroid (NT) cells, assessing the contributions of the nucleus and cytoplasm independently. Our results demonstrate that Raman spectra from the cytoplasmic region can distinguish between FTC and NT cells with an accuracy of 84% under co-culture conditions, maintaining robustness across multiple devices. In contrast, classification based on nuclear spectra achieved only 53% accuracy, suggesting that biochemical alterations in the cytoplasm play a more significant role in FTC detection than those in the nucleus. Our work elevates the promise of Raman-based cytopathology by providing complementary organelle-dependent information to traditional diagnostic methods and demonstrating transferability across different devices.

q-bio.QM↗