Biology subjects

Grouard, S.

Publications and source records attributed to Grouard, S..

4 recordsLinked to original sources

bioRxiv2025

Deciphering Cellular Ecosystems Driving Tumor Progression and Immune Escape from Spatial Transcriptomics and Single-Cell with COMPOTES

Cell-cell communication is central to understanding the complex interactions within the tumor microenvironment. However, current methods fail to identify recurrent communication patterns across patient cohorts from spatial transcriptomics, as they are often limited to single samples or lack essential spatial context. Yet this is essential for understanding how local environments influence cell phenotype and states, and shape the entire cellular ecosystem. We introduce a machine-learning approach that models local, spatially aware ligand-receptor interactions and uses matrix factorization to extract global multicellular programs from large cohorts representing the complex biology of cancer. Applied to a multimodal muscle-invasive bladder cancer cohort of 146 patients, it uncovered 45 communication programs defined by distinct ligand-receptor pairs and cellular niches. In particular, we identified a conserved immune program linked to stalled anti-tumor immunity and a program linking KMT2D loss-of-function mutations with early-stage (T2) tumors, intense proliferation and a favorable response to neoadjuvant chemotherapy.

bioinformatics↗

bioRxiv2025

Multiple instance learning with spatial transcriptomics for interpretable patient-level predictions: application in glioblastoma

Accurate prediction of patient outcomes remains a major challenge in oncology. While recent machine learning (ML) approaches often rely on bulk omics lacking spatial resolution or histology-based multiple instance learning (MIL), spatial transcriptomics (SpT) provides a unique opportunity to capture both molecular content and tissue architecture. However, no generalizable ML framework has yet been established to exploit SpT for patient-level outcomes. We present SpaMIL, a flexible and interpretable MIL framework designed for SpT, with a distillation strategy that enables deployment for hematoxylin and eosin (H&E) slides alone. We evaluate the framework by predicting survival from glioblastoma (GBM) patients, a clinically compelling setting given its aggressiveness with a median survival of only 15 months and the lack of prognostic clinical variables. We analyzed 76 GBM cases from the MOSAIC dataset: 43 with matched SpT, H&E, single-nucleus RNA-seq (scRNA-seq), bulk RNA-seq, and clinical variables, and 33 with H&E for external validation. We developed two main architectures: abMIL, tailored to SpTs spatial molecular structure, and MabMIL, which distills SpT-derived representations into H&E. Model interpretability was achieved through a Shapley-based framework linking prognostic predictions to cell-type compositions via SpT deconvolution. In benchmarking across the five GBM MOSAIC modalities, SpT-based abMIL achieved unprecedented prognostic accuracy (median C-index: 0.72, standard deviation: 0.04), outperforming all other modalities, including established clinical predictors. PCA and deconvolution-based SpT representations surpassed recent foundation models, suggesting the need for further research on SpT foundation models. Our interpretability analysis highlighted malignant and non-malignant cell subpopulations associated with favorable or poor prognosis, consistent with recent reports. Finally, MabMIL maintained strong performance while enabling H&E-only deployment, with improved condorance index over H&E-only baselines in both internal (0.59 vs. 0.57) and external (0.62 vs. 0.55) cohorts.

cancer biology↗

bioRxiv2025

Joint probabilistic modeling of pseudobulk and single-cell transcriptomics enables accurate estimation of cell type composition

Bulk RNA sequencing provides an averaged gene expression profile of the numerous cells in a tissue sample, obscuring critical information about cellular heterogeneity. Computational deconvolution methods can estimate cell type proportions in bulk samples, but current approaches can lack precision in key scenarios due to simplistic statistical assumptions, limited modeling of cell-type heterogeneity and poor handling of rare populations. We present MixupVI, a deep generative model that learns representations of single-cell transcriptomic data and introduces a mixup-based regularization to enable reference-free deconvolution of bulk samples. Our method creates a latent representation with an additive property, where the representation of a pseudobulk sample corresponds to the weighted sum of its constituent cell types. We demonstrate how MixupVI enables accurate estimation of cell type proportions through benchmarking on pseudobulks simulated from a large immune single-cell atlas. To support reproducibility and foster progress in the field, we also release PyDeconv, a Python library that implements multiple state-of-the-art deconvolution algorithms and provides a comprehensive benchmark on simulated pseudobulk datasets.

bioinformatics↗

bioRxiv2024

Robust Evaluation of Deep Learning-based Representation Methods for Survival and Gene Essentiality Prediction on Bulk RNA-seq Data

Deep learning (DL) has shown potential to provide powerful representations of bulk RNA-seq data in cancer research. However, there is no consensus regarding the impact of design choices of DL approaches on the performance of the learned representation, including the model architecture, the training methodology and the various hyperparameters. To address this problem, we evaluate the performance of various design choices of DL representation learning methods using TCGA and DepMap pan-cancer datasets, and assess their predictive power for survival and gene essentiality predictions. We demonstrate that non DL-based baseline methods achieve comparable or superior performance compared to more complex models on survival predictions tasks. DL representation methods, however, are the most efficient to predict the gene essentiality of cell lines. We show that auto-encoders (AE) are consistently improved by techniques such as masking and multi-head training. Our results suggest that the impact of DL representations and of pre-training are highly task- and architecture-dependent, highlighting the need for adopting rigorous evaluation guidelines. These guidelines for robust evaluation are implemented in a pipeline made available to the research community.

bioinformatics↗