Biology subjects

Cantini, L.

Publications and source records attributed to Cantini, L..

At least 19 recordsLinked to original sources

bioRxiv2026

Kontext: Kernel-based method bridging spatial domain identification and cell-cell communication

The spatial organization of biological tissues is characterized by spatial domains: regions with coherent transcriptional and anatomical properties, heavily influenced by local cell-cell interactions. Although identifying spatial domains is a common task in spatial transcriptomics, most existing methods provide limited mechanistic insight into their organization. We present Kontext, a kernel-based framework that jointly represents cells and their local environments, enabling the identification of spatial domains grounded in their underlying ligand-receptor interactions. This representation further provides complementary outputs for characterizing cellular microenvironments, spatial niches, and sender-receiver communication patterns. Kontext outperforms state-of-the-art methods in recovering reference domain annotations across several spatial transcriptomics technologies and biological models, while offering improved computational efficiency and scalability. Applying Kontext on a Xenium temporal mouse heart cryoinjury dataset, we characterize the spatiotemporal evolution of the injury and identify an undescribed sustained activity of the complement system in the communication between fibroblasts and macrophages during scar maturation. In a million-cell three-dimensional Open-ST dataset of human metastatic lymph node tissue, we combine spatial domain identification and niche analysis to define the boundary region between the tumor and its microenvironment, revealing distinct communication patterns between this interface and the tumor core.

bioinformatics↗

bioRxiv2026

Linking transcriptome to cell behavior in real time uncovers molecular fate-asymmetry and a quiescence cycle among adult neural stem cells

Adult neural stem cells (NSCs) are transcriptionally heterogeneous, yet the relationship between molecular heterogeneities and individual NSC trajectories remains unclear. Here, we link transcriptional identities with cellular features and measures of real time to identify molecular trajectories hidden within transcriptomic space. Among self-renewing NSCs, we resolve single-cell molecular transitions associated with fate asymmetry at division. We also show that individual self-renewing NSCs progressively transition from molecular states of deep to shallow quiescence during prolonged quiescence phases. Together, this work reveals individual NSC trajectories within transcriptomic space during fate decisions and state transitions. In particular, it highlights the existence of a transcriptionally encoded quiescence cycle followed by adult NSCs, independent of lineage progression, that balances division with cell growth to sustain self-renewal over time.

developmental biology↗

bioRxiv2026

Inferring Cell Fate Trajectories in Time-Resolved Metabolic RNA Labeling data

Single-cell RNA sequencing provides high-resolution snapshots of cellular states but lacks direct information about transcriptional dynamics. Metabolic RNA labeling addresses this limitation by distinguishing newly synthesized RNA, offering insight into the direction of cell state changes, and providing valuable information when attempting to recover the underlying continuous dynamics from static snapshots of cell distributions. However, existing trajectory inference methods do not fully exploit this additional signal. Here, we propose FLOWSATATE, a framework for single-cell trajectory inference that leverages time-resolved RNA labeling within an Optimal Transport setting. We model cell dynamics as a gradient flow in an inferred potential landscape parameterized by a neural network, integrating both total and labeled RNA across time points. The learned potential enables identification of key genes and transcription factors driving cell fate decisions and supports prediction of future cellular states. We benchmark our approach on its ability to generalize unseen data and recover coherent trajectories. We also apply it to study colorectal cancer response to demethylation treatment as well as neuronal differentiation of embryonic stem cells.

bioinformatics↗

bioRxiv2026

CHAMPOLLION: Robust Multi-Omics Integration via Inverse Optimal Transport Using Paired Cells

Fully capturing cellular identity requires integrating multiple molecular layers. Bridge integration, i.e. aligning unimodal datasets using a paired multi-omic reference, has emerged as a practical solution, yet existing methods offer limited interpretability and use paired information without regularization, making them sensitive to limited size and coverage. We introduce CHAMPOLLION, which uses regularized optimal transport to learn an interpretable cross-modal metric that drives the alignment of unpaired cells while capturing relationships between molecular features. Benchmarks on RNA-protein and RNA-ATAC datasets show that CHAMPOLLION outperforms existing approaches, remaining accurate with few paired cells and even generalizing to unseen cell types. Beyond alignment, CHAMPOLLION reveals biologically meaningful cross-modal relationships, highlighting in scRNA-protein data a potential role for CD18 across multiple cancers, and, in a human tonsil atlas combining scRNA-seq and scATAC-seq, suggesting that MEF2C may regulate inflammatory responses beyond the brain, notably in plasmacytoid dendritic cells.

bioinformatics↗

bioRxiv2026

Modelling multicellular coordination by bridging cell-cell communication and intracellular regulation through multilayer networks

In multicellular organisms, cells with various roles and locations coordinate to provide systemic and cohesive response to perturbations. These complex behaviors emerge from a complex interplay between intracellular regulation and intercellular signals that mediate cell-cell communication. While single-cell technologies opened the possibility of studying both, most methods focus solely on one of these aspects. Thus, they are only able to partially recover in vivo and multicellular behaviors. We here introduce ReCoN (REconstruction of multicellular COordination Networks from single-cell data), a framework combining intracellular gene regulation and cell-cell communication to provide insights into multicellular coordination from single-cell data. First, ReCoN infers from single-cell data a heterogeneous multilayer network containing both cell-type-specific intracellular subnetworks and ligand-receptor interactions. Through random walk with restart explorations, ReCoN then infers the response of each cell type to both intra- and extracellular perturbations, such as a gene knock-out or a cytokine, respectively. ReCoN was evaluated on predicting the in vivo response of immune cell-types to different cytokines and on recovering cardiac cell-type response in heart failure. It highlighted the role of indirect effects, where cells emit secondary messengers in response to the initial perturbation to coordinate multicellular transcriptomic responses. Additionally, ReCoN predicted distinct fibroblast states emerging in different microenvironments reconstructed from spatial data. ReCoN provides an interpretable modeling framework for multicellular systems that allows for the simulation of perturbations, including the assessment of the cellular selectivity of these treatments in vivo. Ultimately, it can help design patient-specific molecular therapies. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=107 SRC="FIGDIR/small/700561v1_ufig1.gif" ALT="Figure 1"> View larger version (34K): org.highwire.dtl.DTLVardef@7cdceforg.highwire.dtl.DTLVardef@18cfa3eorg.highwire.dtl.DTLVardef@812645org.highwire.dtl.DTLVardef@f7bce5_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2025

scPRINT-2: Towards the next-generation of cell foundation models and benchmarks

Cell biology has been booming with foundation models trained on large single-cell RNA-seq databases, but benchmarks and capabilities remain unclear. We propose an additive benchmark across a gymnasium of tasks to discover which features improve performance. From these findings, we present scPRINT-2, a single-cell Foundation Model pre-trained across 350 million cells and 16 organisms. Our contributions in pre-training tasks, tokenization, and losses made scPRINT-2 state-of-the-art in expression denoising, cell embedding, and cell type prediction. Furthermore, with our cell-level architecture, scPRINT-2 becomes generative, as demonstrated by our expression imputation and counterfactual reasoning results. Finally, thanks to our pre-training database, we uncover generalization to unseen modalities and organisms. These studies, together with improved abilities in gene embeddings and gene network inference, place scPRINT-2 as a next-generation cell foundation model.

bioinformatics↗

bioRxiv2025

CIRCE: a scalable Python package to predict cis-regulatory DNA interactions from single-cell chromatin accessibility data

Chromatin 3D folding creates numerous DNA interactions, participating in gene expression regulation. Single-cell chromatin-accessibility assays now profile hundreds of thousands of cells, challenging existing methods for mapping cis-regulatory interactions. We present CIRCE, a fast and scalable Python package to predict cis-regulatory DNA interactions from single-cell chromatin accessibility data. CIRCE re-implements the Cicero workflow to analyse single-cell atlases, cutting runtime and memory use by several orders of magnitude. We also provide new options to compute metacells, grouping similar cells to reduce data sparsity. We benchmarked CIRCE against Cicero on two datasets of different sizes and demonstrated the improvement from CIRCEs metacells strategy with promoter capture Hi-C data. We also evaluated how DNA interaction predictions are impacted by different pre-processing. We observed a negative impact of Ciceros count normalization, and the best performance was obtained with the single-cell count matrix directly. Finally, we demonstrated the scalability of CIRCE by processing a dataset of more than 700000 cells and 1 million DNA regions in less than an hour. CIRCE should greatly facilitate the prediction of DNA region interactions for scverse and Python users, while providing new and up-to-date pre-processing insights. Availability and reproducibilityCIRCE is released as an open-source software under the AGPL-3.0 license. The package source code is available on GitHub at https://github.com/cantinilab/CIRCE, and its documentation is accessible at https://circe.readthedocs.io. The code to reproduce the presented results is available as a Snakemake pipeline at https://github.com/cantinilab/circe_reproducibility.

bioinformatics↗

bioRxiv2025

Non-centromeric CENP-A epigenetically regulates epithelial-mesenchymal plasticity and heterogeneity in human cells

The centromeric histone variant CENP-A, highly expressed in aggressive cancers, can promote an epithelial-mesenchymal transition (EMT) whose underlying mechanisms remain to decipher. Here, we tracked both the temporal dynamics of EMT states and CENP-A localization over time using a reversible high CENP-A expression system in human cells. Cell populations presenting hybrid EMT states at start, when exposed to high CENP-A levels, progressively accumulated mesenchymal states and displayed increased centromeric and ectopic CENP-A incorporation. Mechanistically, we reveal ectopic CENP-A gains at EMT genes by chromatin immunoprecipitation and identify two distinct EMT programs activated at different stages of the cell cycle by single-nucleus multi-omics. Importantly, while a pre-existent inflammatory program got amplified, high CENP-A induced a new developmental program. Remarkably, interrupting high CENP-A provision erased induced programs along with ectopic CENP-A incorporation, in line with non-genetic alterations. Our findings uncover an unconventional, non-centromeric function for CENP-A in epigenetically modulating epithelial-mesenchymal plasticity.

molecular biology↗

bioRxiv2025

Towards foundation models that learn across biological scales

We have reached a point where many bio foundation models exist across 4 different scales, from molecules to molecular chains, cells, and tissues. However, while related in many ways, these models do not yet bridge these scales. We present a framework and architecture called Xpressor that enables cross-scale learning by (1) using a novel cross-attention mechanism to compress high-dimensional gene representations into lower-dimensional cell-state vectors, and (2) implementing a multi-scale fine-tuning approach that allows cell models to leverage and adapt protein-level representations. Using a cell Foundation Model as an example, we demonstrate that our architecture improves model performance across multiple tasks, including cell-type prediction (+12%) and embedding quality (+8%). Together, these advances represent first steps toward models that can understad and bridge different scales of biological organization.

bioinformatics↗

bioRxiv2024

IL-23 tunes inflammatory functions of human mucosal-associated invariant T (MAIT) cells

IL-23 signaling plays a key role in the pathogenesis of chronic inflammatory and infectious diseases, yet the cellular targets and signaling pathways affected by this cytokine remain poorly understood. We show that IL-23 receptors are expressed on the large majority of human MAIT, but not of conventional T cells. Protein and transcriptional profiling at the population and single cell level demonstrates that stimulation with IL-23 or the structurally related cytokine IL-12 drives distinct functional profiles, revealing a high level of plasticity of MAIT cells. IL-23, in particular, affects key molecules and pathways related to autoimmunity and cytotoxic functions. Integrated analysis of transcriptomic and chromatin accessibility, supported by CRISPR/Cas9 mediated deletion, shows that AP-1 transcription factors constitute a key regulatory node of the IL-23 pathway in MAIT cells. In conclusion, our findings indicate that MAIT cells are key mediators of IL-23 functions in immunity to infections and chronic inflammatory diseases.

immunology↗

bioRxiv2024

scPRINT: pre-training on 50 million cells allows robust gene network predictions

A cell is governed by the interaction of myriads of macromolecules. Such a network of interaction has remained an elusive milestone in cellular biology. Building on recent advances in large foundation models and their ability to learn without supervision, we present scPRINT, a large cell model for the inference of gene networks pre-trained on more than 50M cells from the cellxgene database. Using novel pretraining methods and model architecture, scPRINT pushes large transformer models towards more interpretability and usability in uncovering the complex biology of the cell. Based on our atlas-level benchmarks, scPRINT demonstrates superior performance in gene network inference to the state of the art, as well as competitive zero-shot abilities in denoising, batch effect correction, and cell label prediction. On an atlas of benign prostatic hyperplasia, scPRINT highlights the profound connections between ion exchange, senescence, and chronic inflammation.

bioinformatics↗

bioRxiv2024

Learning cell fate landscapes from spatial transcriptomics using Fused Gromov-Wasserstein

In dynamic biological processes such as development, spatial transcriptomics is revolutionizing the study of the mechanisms underlying spatial organization within tissues. Inferring cell fate trajectories from spatial transcriptomics profiled at several time points has thus emerged as a critical goal, requiring novel computational methods. Wasserstein gradient flow learning is a promising framework for analyzing sequencing data across time, built around a neural network representing the differentiation potential. However, existing gradient flow learning methods cannot analyze spatially resolved transcriptomic data. Here, we propose STORIES, a method that employs an extension of Optimal Transport to learn a spatially informed potential. We benchmark our approach using three large Stereo-seq spatiotemporal atlases and demonstrate superior spatial coherence compared to existing approaches. Finally, we provide an in-depth analysis of axolotl neural regeneration and mouse gliogenesis, recovering gene trends for known markers as Nptx1 in neuron regeneration and Aldh1l1 in gliogenesis and additional putative drivers.

bioinformatics↗

bioRxiv2024

MOTL: enhancing multi-omics matrix factorization with transfer learning

Joint matrix factorization is popular for extracting lower dimensional representations of multi-omics data but loses effectiveness with limited samples. Addressing this limitation, we introduce MOTL (Multi-Omics Transfer Learning), a framework that enhances MOFA (Multi-Omics Factor Analysis) by inferring latent factors for small multi-omics target datasets with respect to those inferred from a large heterogeneous learning dataset. We evaluated MOTL by designing simulated and real data protocols and demonstrated that MOTL improves the factorization of limited-sample multi-omics datasets when compared to factorization without transfer learning. When applied to actual glioblastoma samples, MOTL enhanced delineation of cancer status and subtype.

bioinformatics↗

bioRxiv2024

scConfluence : single-cell diagonal integration with regularized Inverse Optimal Transport on weakly connected features

The abundance of unpaired multimodal single-cell data has motivated a growing body of research into the development of diagonal integration methods. However, the state-of-the-art suffers from the loss of biological information due to feature conversion and struggles with modality-specific populations. To overcome these crucial limitations, we here introduced scConfluence, a novel method for single-cell diagonal integration. scConfluence combines uncoupled autoencoders on the complete set of features with regularized Inverse Optimal Transport on weakly connected features. We extensively benchmark scConfluence in several single-cell integration scenarios proving that it outperforms the state-of-the-art. We then demonstrate the biological relevance of scConfluence in three applications. We predict spatial patterns for Scgn, Synpr and Olah in scRNA-smFISH integration. We improve the classification of B cells and Monocytes in highly heterogeneous scRNA-scATAC-CyTOF integration. Finally, we reveal the joint contribution of Fezf2 and apical dendrite morphology in Intra Telencephalic neurons, based on morphological images and scRNA.

bioinformatics↗

bioRxiv2023

Molecular mechanisms reconstruction from single-cell multi-omics data with HuMMuS

The molecular identity of a cell results from a complex interplay between heterogeneous molecular layers. Recent advances in single-cell sequencing technologies have opened the possibility to measure such molecular layers of regulation. Here, we present HuMMuS, a new method for inferring regulatory mechanisms from single-cell multi-omics data. Differently from the state-of-the-art, HuMMuS captures cooperation between biological macromolecules and can easily include additional layers of molecular regulation. We benchmarked HuMMuS with respect to the state-of-the-art on both paired and unpaired multi-omics datasets. Our results proved the improvements provided by HuMMus in terms of TF targets, TF binding motifs and regulatory regions prediction. Finally, once applied to snmC-seq, scATAC-seq and scRNA-seq data from mouse brain cortex, HuMMuS enabled to accurately cluster scRNA profiles and to identify potential driver TFs.

bioinformatics↗

bioRxiv2023

Paired single-cell multi-omics data integration with Mowgli

The profiling of multiple molecular layers from the same set of cells has recently become possible. There is thus a growing need for multi-view learning methods able to jointly analyze these data. We here present Multi-Omics Wasserstein inteGrative anaLysIs (Mowgli), a novel method for the integration of paired multi-omics data with any type and number of omics. Of note, Mowgli combines integrative Nonnegative Matrix Factorization (NMF) and Optimal Transport (OT), enhancing at the same time the clustering performance and interpretability of integrative NMF. We apply Mowgli to multiple paired single-cell multi-omics data profiled with 10X Multiome, CITE-seq and TEA-seq. Our in depth benchmark demonstrates that Mowglis performance is competitive with the state-of-the-art in cell clustering and superior to the state-of-the-art once considering biological interpretability. Mowgli is implemented as a Python package seamlessly integrated within the scverse ecosystem and it is available at http://github.com/cantinilab/mowgli.

bioinformatics↗

bioRxiv2021

Optimal Transport improves cell-cell similarity inference in single-cell omics data

The recent advent of high-throughput single-cell molecular profiling is revolutionizing biology and medicine by unveiling the diversity of cell types and states contributing to development and disease. The identification and characterization of cellular heterogeneity is typically achieved through unsupervised clustering, which crucially relies on a similarity metric. We here propose the use of Optimal Transport (OT) as a cell-cell similarity metric for single-cell omics data. OT defines distances to compare, in a geometrically faithful way, high-dimensional data represented as probability distributions. It is thus expected to better capture complex relationships between features and produce a performance improvement over state-of-the-art metrics. To speed up computations and cope with the high-dimensionality of single-cell data, we consider the entropic regularization of the classical OT distance. We then extensively benchmark OT against state-of-the-art metrics over thirteen independent datasets, including simulated, scRNA-seq, scATAC-seq and single-cell DNA methylation data. First, we test the ability of the metrics to detect the similarity between cells belonging to the same groups (e.g. cell types, cell lines of origin). Then, we apply unsupervised clustering and test the quality of the resulting clusters. In our in-depth evaluation, OT is found to improve cell-cell similarity inference and cell clustering in all simulated and real scRNA-seq data, while its performances are comparable with Pearson correlation in scATAC-seq and single-cell DNA methylation data. All our analyses are reproducible through the OT-scOmics Jupyter notebook available at https://github.com/ComputationalSystemsBiology/OT-scOmics.

systems biology↗

bioRxiv2020

Evaluating the reproducibility of single-cell gene regulatory network inference algorithms

Networks are powerful tools to represent and investigate biological systems. The development of algorithms inferring regulatory interactions from functional genomics data has been an active area of research. With the advent of single-cell RNA-seq data (scRNA-seq), numerous methods specifically designed to take advantage of single-cell datasets have been proposed. However, published benchmarks on single-cell network inference are mostly based on simulated data. Once applied to real data, these benchmarks take into account only a small set of genes and only compare the inferred networks with an imposed ground-truth. Here, we benchmark four single-cell network inference methods based on their reproducibility, i.e. their ability to infer similar networks when applied to two independent datasets for the same biological condition. We tested each of these methods on real data from three biological conditions: human retina, T-cells in colorectal cancer, and human hematopoiesis. GENIE3 results to be the most reproducible algorithm, independently from the single-cell sequencing platform, the cell type annotation system, the number of cells constituting the dataset, or the thresholding applied to the links of the inferred networks. In order to ensure the reproducibility and ease extensions of this benchmark study, we implemented all the analyses in scNET, a Jupyter notebook available at https://github.com/ComputationalSystemsBiology/scNET.

systems biology↗