Vincent Cabeli
Vincent Cabeli has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Hervé Isambert
CNRS UMR168, Institut Curie, Université PSL, Sorbonne Université
2 shared publications - Jean Ogier du Terrail
Living Models, Paris, France
2 shared publications Léonard Strouk
Living Models, Paris, France
2 shared publicationsNadir Sella
Institut Roche, Boulogne-Billancourt, France
2 shared publications- Alexey D. Rukhovich
V.V. Dokuchaev Soil Science Institute, Pyzhevskiy Pereulok 7, 119017 Moscow, Russia
1 shared publication - Amélie Barozet
Living Models, Paris, France
1 shared publication
Latest Publications
- Open Access
BOTANIC-1: a series of long-context plant genomic foundation models in the agentic era
The development of climate-resilient crops would be greatly accelerated by models able to reason directly over plant genomic sequences and to pinpoint trait-associated regions or loci. Anticipating the impact of DNA base changes (variants) remains challenging, and understanding regulatory mechanisms is still an active area of research. Through self-supervised training on unannotated genomic data, genomic language models (gLMs) can learn DNA syntax and grammar that go beyond current annotations, thus complementing standard bioinformatics analyses that rely on rules established by decades of genomics research. Here we present our agent-powered Model Factory and its first outputs: the Botanic1 family of gLMs designed for plant research, which operates reliably on sequences from hundreds of base pairs up to 128 kbp. These models outperform all generalist and plant-specific gLMs (as well as specialised baselines) on one of the largest sets of plant genomics evaluation tasks reported to date, at a much smaller budget than concurrent models. Mechanistic interpretability analysis identifies features associated with biologically meaningful sequence properties including coding region boundaries and splice site motifs, demonstrating that these models are a source of biological insight beyond their benchmark performance. Finally, because a gLM only becomes practically useful when embedded in a broader workflow, we integrate Botanic1 as a specialised genomic layer callable by a generalist large language model (LLM) agent, illustrating how such hybrid systems could accelerate plant biology research. To support the plant genomics research community, we release the four Botanic1 models, their pre-training corpus and the trained sparse autoencoder for research use at https://huggingface.co/spaces/living-models/botanic1-report .
Open AccessBOTANIC-1: a series of long-context plant genomic foundation models in the agentic era
The development of climate-resilient crops would be greatly accelerated by models able to reason directly over plant genomic sequences and to pinpoint trait-associated regions or loci. Anticipating the impact of DNA base changes (variants) remains challenging, and understanding regulatory mechanisms is still an active area of research. Through self-supervised training on unannotated genomic data, genomic language models (gLMs) can learn DNA syntax and grammar that go beyond current annotations, thus complementing standard bioinformatics analyses that rely on rules established by decades of genomics research. Here we present our agent-powered Model Factory and its first outputs: the Botanic1 family of gLMs designed for plant research, which operates reliably on sequences from hundreds of base pairs up to 128 kbp. These models outperform all generalist and plant-specific gLMs (as well as specialised baselines) on one of the largest sets of plant genomics evaluation tasks reported to date, at a much smaller budget than concurrent models. Mechanistic interpretability analysis identifies features associated with biologically meaningful sequence properties including coding region boundaries and splice site motifs, demonstrating that these models are a source of biological insight beyond their benchmark performance. Finally, because a gLM only becomes practically useful when embedded in a broader workflow, we integrate Botanic1 as a specialised genomic layer callable by a generalist large language model (LLM) agent, illustrating how such hybrid systems could accelerate plant biology research. To support the plant genomics research community, we release the four Botanic1 models, their pre-training corpus and the trained sparse autoencoder for research use at https://huggingface.co/spaces/living-models/botanic1-report .
- 4 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Learning causal graphs from continuous or mixed datasets of biological or clinical interest
Apprentissage de graphes causaux à partir de données continues ou mixtes d’intérêt biologique ou clinique: Les travaux de cette thèse s’inscrivent dans la théorie principalement développée par Judea Pearl sur les diagrammes causaux; des modèles graphiques qui permettent de dériver toutes les quantités causales d’intérêt formellement et intuitivement. Nous traitons le problème de l’inférence de réseau causal à partir uniquement de données d’observation, c’est-à-dire sans aucune intervention de la part de l’expérimentateur. En particulier, nous proposons d’améliorer les méthodes existantes pour les rendre plus aptes à analyser des données issues du monde réel, en nous affranchissant le plus possible des contraintes sur les distributions des données, et en les rendant plus interprétables. Nous proposons une extension de MIIC, une approche basée sur les contraintes et la théorie de l’information pour retrouver la classe d’équivalence du graphe causal à partir d’observations. Notre contribution est un algorithme de discrétisation optimale basé sur le principe de description minimale pour simultanément estimer la valeur de l’information mutuelle (et multivariée) et évaluer sa significativité entre des échantillons de variables de n’importe quelle nature : continue, catégorique ou mixte. Nous mettons à profit ces développements pour analyser des jeux de données mixtes d'intérêt clinique (dossiers médicaux de patients atteints de troubles cognitifs; ou du cancer du sein) ou biologique (réseaux de régulation génique de cellules précurseur hématopoïétiques).
- 4 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Learning causal graphs from continuous or mixed datasets of biological or clinical interest
Apprentissage de graphes causaux à partir de données continues ou mixtes d’intérêt biologique ou clinique: Les travaux de cette thèse s’inscrivent dans la théorie principalement développée par Judea Pearl sur les diagrammes causaux; des modèles graphiques qui permettent de dériver toutes les quantités causales d’intérêt formellement et intuitivement. Nous traitons le problème de l’inférence de réseau causal à partir uniquement de données d’observation, c’est-à-dire sans aucune intervention de la part de l’expérimentateur. En particulier, nous proposons d’améliorer les méthodes existantes pour les rendre plus aptes à analyser des données issues du monde réel, en nous affranchissant le plus possible des contraintes sur les distributions des données, et en les rendant plus interprétables. Nous proposons une extension de MIIC, une approche basée sur les contraintes et la théorie de l’information pour retrouver la classe d’équivalence du graphe causal à partir d’observations. Notre contribution est un algorithme de discrétisation optimale basé sur le principe de description minimale pour simultanément estimer la valeur de l’information mutuelle (et multivariée) et évaluer sa significativité entre des échantillons de variables de n’importe quelle nature : continue, catégorique ou mixte. Nous mettons à profit ces développements pour analyser des jeux de données mixtes d'intérêt clinique (dossiers médicaux de patients atteints de troubles cognitifs; ou du cancer du sein) ou biologique (réseaux de régulation génique de cellules précurseur hématopoïétiques).
- Open Access
BOTANIC-0: a series of foundation models for plant genomic data
Genomic language models (gLMs) have emerged as a powerful paradigm for learning regulatory biology directly from DNA sequence. Here, we introduce Botanic0 ¶ , a family of plant genomic foundation models spanning 100M to 1B parameters and pretrained on 43 phylogenetically diverse plant genomes. The Botanic0-S , Botanic0-M , and Botanic0-L models form the first generation of a long-term research initiative, dedicated to advancing crop improvement research, genotype–to-phenotype modeling, and sequence-based genome editing. The architecture, pre-training pipeline and pre-training dataset of Botanic0 follow the seminal work of [1]. Across a broad suite of genomic and genetic prediction tasks, including regulatory element annotation, gene expression inference, and variant effect prediction, Botanic0 models achieve performance competitive with state-of-the-art foundation models, both in zero-shot settings and after fine-tuning. Scaling analyses reveal consistent improvements in predictive power with increased model capacity, highlighting the benefits of large-model pretraining for plant genomics. This work establishes our ability to train foundation models at scale, and lays the foundation for the next generations of models to come. To support reproducible research and community benchmarking, we release all Botanic0 models at https://huggingface.co/living-models/models .
Open AccessBOTANIC-0: a series of foundation models for plant genomic data
Genomic language models (gLMs) have emerged as a powerful paradigm for learning regulatory biology directly from DNA sequence. Here, we introduce Botanic0 ¶ , a family of plant genomic foundation models spanning 100M to 1B parameters and pretrained on 43 phylogenetically diverse plant genomes. The Botanic0-S , Botanic0-M , and Botanic0-L models form the first generation of a long-term research initiative, dedicated to advancing crop improvement research, genotype–to-phenotype modeling, and sequence-based genome editing. The architecture, pre-training pipeline and pre-training dataset of Botanic0 follow the seminal work of [1]. Across a broad suite of genomic and genetic prediction tasks, including regulatory element annotation, gene expression inference, and variant effect prediction, Botanic0 models achieve performance competitive with state-of-the-art foundation models, both in zero-shot settings and after fine-tuning. Scaling analyses reveal consistent improvements in predictive power with increased model capacity, highlighting the benefits of large-model pretraining for plant genomics. This work establishes our ability to train foundation models at scale, and lays the foundation for the next generations of models to come. To support reproducible research and community benchmarking, we release all Botanic0 models at https://huggingface.co/living-models/models .
- Open Access
- 10 August 2022
- Published by Springer Nature in npj Digital Medicine
- Vol. 5 (1) , 1-10
Interactive exploration of a global clinical network from a large breast cancer cohort
Despite unprecedented amount of information now available in medical records, health data remain underexploited due to their heterogeneity and complexity. Simple charts and hypothesis-driven statistics can no longer apprehend the content of information-rich clinical data. There is, therefore, a clear need for powerful interactive visualization tools enabling medical practitioners to perceive the patterns and insights gained by state-of-the-art machine learning algorithms. Here, we report an interactive graphical interface for use as the front end of a machine learning causal inference server (MIIC), to facilitate the visualization and comprehension by clinicians of relationships between clinically relevant variables. The widespread use of such tools, facilitating the interactive exploration of datasets, is crucial both for data visualization and for the generation of research hypotheses. We demonstrate the utility of the MIIC interactive interface, by exploring the clinical network of a large cohort of breast cancer patients treated with neoadjuvant chemotherapy (NAC). This example highlights, in particular, the direct and indirect links between post-NAC clinical responses and patient survival. The MIIC interactive graphical interface has the potential to help clinicians identify actionable nodes and edges in clinical networks, thereby ultimately improving the patient care pathway.
- 10 August 2022
- Published by Springer Nature in npj Digital Medicine
- Vol. 5 (1) , 1-10
Open AccessInteractive exploration of a global clinical network from a large breast cancer cohort
Despite unprecedented amount of information now available in medical records, health data remain underexploited due to their heterogeneity and complexity. Simple charts and hypothesis-driven statistics can no longer apprehend the content of information-rich clinical data. There is, therefore, a clear need for powerful interactive visualization tools enabling medical practitioners to perceive the patterns and insights gained by state-of-the-art machine learning algorithms. Here, we report an interactive graphical interface for use as the front end of a machine learning causal inference server (MIIC), to facilitate the visualization and comprehension by clinicians of relationships between clinically relevant variables. The widespread use of such tools, facilitating the interactive exploration of datasets, is crucial both for data visualization and for the generation of research hypotheses. We demonstrate the utility of the MIIC interactive interface, by exploring the clinical network of a large cohort of breast cancer patients treated with neoadjuvant chemotherapy (NAC). This example highlights, in particular, the direct and indirect links between post-NAC clinical responses and patient survival. The MIIC interactive graphical interface has the potential to help clinicians identify actionable nodes and edges in clinical networks, thereby ultimately improving the patient care pathway.
- Open Access
Inferring Gene Networks in Bone Marrow Hematopoietic Stem Cell-Supporting Stromal Niche Populations
- Open Access
- 18 May 2020
- Published by Public Library of Science (PLoS) in PLOS Computational Biology
- Vol. 16 (5) , e1007866
Learning clinical networks from medical records based on information estimates in mixed-type data
The precise diagnostics of complex diseases require to integrate a large amount of information from heterogeneous clinical and biomedical data, whose direct and indirect interdependences are notoriously difficult to assess. To this end, we propose an efficient computational approach to simultaneously compute and assess the significance of multivariate information between any combination of mixed-type (continuous/categorical) variables. The method is then used to uncover direct, indirect and possibly causal relationships between mixed-type data from medical records, by extending a recent machine learning method to reconstruct graphical models beyond simple categorical datasets. The method is shown to outperform existing tools on benchmark mixed-type datasets, before being applied to analyze the medical records of eldery patients with cognitive disorders from La Pitié-Salpêtrière Hospital, Paris. The resulting clinical network visually captures the global interdependences in these medical records and some facets of clinical diagnosis practice, without specific hypothesis nor prior knowledge on any clinically relevant information. In particular, it provides some physiological insights linking the consequence of cerebrovascular accidents to the atrophy of important brain structures associated to cognitive impairment. We developed a machine learning approach to analyze medical records and help clinicians visualize the direct and indirect interrelations between clinical examinations and the variety of syndromes implicated in complex diseases. The reconstruction of such clinical networks is illustrated on the spectrum of cognitive disorders, originating from either neurodegenerative, cerebrovascular or psychiatric dementias. This global network analysis is also shown to uncover novel direct associations and possible cause-effect relationships between clinically relevant information, such as medical examinations, diagnoses, treatments and personal data from patients’ medical records.
- 18 May 2020
- Published by Public Library of Science (PLoS) in PLOS Computational Biology
- Vol. 16 (5) , e1007866
Open AccessLearning clinical networks from medical records based on information estimates in mixed-type data
The precise diagnostics of complex diseases require to integrate a large amount of information from heterogeneous clinical and biomedical data, whose direct and indirect interdependences are notoriously difficult to assess. To this end, we propose an efficient computational approach to simultaneously compute and assess the significance of multivariate information between any combination of mixed-type (continuous/categorical) variables. The method is then used to uncover direct, indirect and possibly causal relationships between mixed-type data from medical records, by extending a recent machine learning method to reconstruct graphical models beyond simple categorical datasets. The method is shown to outperform existing tools on benchmark mixed-type datasets, before being applied to analyze the medical records of eldery patients with cognitive disorders from La Pitié-Salpêtrière Hospital, Paris. The resulting clinical network visually captures the global interdependences in these medical records and some facets of clinical diagnosis practice, without specific hypothesis nor prior knowledge on any clinically relevant information. In particular, it provides some physiological insights linking the consequence of cerebrovascular accidents to the atrophy of important brain structures associated to cognitive impairment. We developed a machine learning approach to analyze medical records and help clinicians visualize the direct and indirect interrelations between clinical examinations and the variety of syndromes implicated in complex diseases. The reconstruction of such clinical networks is illustrated on the spectrum of cognitive disorders, originating from either neurodegenerative, cerebrovascular or psychiatric dementias. This global network analysis is also shown to uncover novel direct associations and possible cause-effect relationships between clinically relevant information, such as medical examinations, diagnoses, treatments and personal data from patients’ medical records.