Search

International audience; Plant genomes are characterized by large and complex gene families that often result in similar and partially overlapping functions 1. This genetic redundancy severely hampers current efforts to uncover novel phenotypes, delaying basic genetic research and breeding programs 2. Here, we describe the development and validation of Multi-Knock, a genome-scale CRISPR toolbox that overcomes functional redundancy in Arabidopsis by simultaneously targeting multiple gene-family members, thus identifying genetically hidden components. We computationally designed 59,129 optimal single guide RNAs (sgRNAs) that each target 2 to 10 genes within a family at once. Furthermore, partitioning the library into ten sub-libraries directed towards a different functional group allows flexible and targeted genetic screens. From the 5,635 sgRNAs targeting the plant transportome, we generated over 3,500 independent Arabidopsis lines that allowed us to identify and characterize the first known cytokinin tonoplast-localized transporters in plants. With the ability to overcome functional redundancy in plants at the genome-scale level, the developed strategy can be readily deployed by scientists and breeders for basic research and to expedite breeding efforts.

International audience

International audience; To benefit from virtual reality (VR) as a complementary tool for training, coaches must determine the proper tools and variables for tracking sports performance. We explored the basketball shooting at several scales (basket-ball, ball-player, and player systems) by monitoring success-rate, and ball and body kinematics. We measured how these scales of analysis allowed tracking players' expertise and perceptual sensitivity to basket distance. Experienced and novice players were instructed to naturally throw and swish an instrumented ball in a stereoscopically rendered virtual basket. We challenged their perceptual-motor systems by manipulating the distance of the virtual basket while keeping the surrounding environment unchanged. The success-rate accounted for the players' shooting adjustments to the manipulation of basket distance and allowed tracking their expertise. Ball kinematics also reflected the manipulation of distance and allowed detecting gender, but did not reflect the players' expertise. Finally, body kinematics variables did not echo players' adjustments to the distance manipulation but reflected their expertise and gender. The results gained at each scale of analysis are discussed with regard to the simulator's construct, biomechanical, and psychological fidelity.

Design Thyroid-stimulating hormone deficiency (TSHD) is a rare disease. It may be isolated, secondary to abnormalities of genes involved in TSH biosynthesis, or associated with other pituitary deficits or abnormalities of genes involved in pituitary ontogenesis. Several genes are involved in thyrotroph development and function. Objective Our aim was to determine the genetic causes of TSHD, either isolated (ITSHD) or associated with somatotroph deficiency (TSHD-GHD), in the cohort of patients from the GENHYPOPIT network. Methods Next-generation sequencing (NGS) analyses were performed as a panel of genes on a cohort of patients with non-syndromic ITSHD or TSHGHD. The variants were classified according to the American College of Medical Genetics classification reviewed by the NGS-Diag network and correlated with the phenotype. Class 3, 4, and 5 single-nucleotide variants were checked by Sanger sequencing and copy number variants by multiplex ligation-dependent probe amplification (MLPA). Results A total of 64 index cases (22 ITSHD and 42 TSHD-GHD) were included in this cohort. A genetic cause was identified in 26.5% of patients, with 36.3% in the ITSHD group (variants in TSHβ and IGSF1) and 21.4% in TSHD-GHD (variants in IGSF1, TSHβ, TRHR, GH1, POU1F1, and PROP1). Among the pathogenic and likely pathogenic variants identified, 42% were in IGSF1, including six not previously reported. Conclusion Our results show that IGSF1 variants represent the most frequent aetiology of TSH deficiency. Despite a systematic NGS approach and the identification of new variants, most patients remain without a molecular diagnosis. Larger scale studies, such as exome or genome studies, should be considered in the future.

Fluorescent proteins (FPs) are a powerful tool to quantitatively monitor gene expression. The dynamics of a promoter and its regulation can be inferred from fluorescence data. The interpretation of fluorescent data, however, is strongly dependent on the maturation of FPs since different proteins mature in distinct ways. We propose a novel approach for analyzing fluorescent reporter data by incorporating maturation dynamics in the reconstruction of promoter activities. Our approach consists of developing and calibrating mechanistic maturation models for distinct FPs. These models are then used alongside a Bayesian approach to estimate promoter activities from fluorescence data. We demonstrate by means of targeted experiments in Escherichia coli that our approach provides robust estimates and that accounting for maturation is, in many cases, essential for the interpretation of gene expression data.

MotivationInteractions between proteins help us understand how genes are functionally related and how they contribute to phenotypes. Experiments provide imperfect ‘ground truth’ information about a small subset of potential interactions in a specific biological context, which can then be extended to the whole genome across different contexts, such as conditions, tissues or species, through machine learning methods. However, evaluating the performance of these methods remains a critical challenge. Here, we propose to evaluate the generalizability of gene characterizations through the shape of performance curves.ResultsWe identify Functional Equivalence Classes (FECs), subsets of annotated and unannotated genes that jointly drive performance, by assessing the presence of straight lines in ROC curves built from gene-centric prediction tasks, such as function or interaction predictions. FECs are widespread across data types and methods, they can be used to evaluate the extent and context-specificity of functional annotations in a data-driven manner. For example, FECs suggest that B cell markers can be decomposed into shared primary markers (10–50 genes), and tissue-specific secondary markers (100–500 genes). In addition, FECs suggest the existence of functional modules that span a wide range of the genome, with marker sets spanning at most 5% of the genome and data-driven extensions of Gene Ontology sets spanning up to 40% of the genome. Simple to assess visually and statistically, the identification of FECs in performance curves paves the way for novel functional characterization and increased robustness in the definition of functional gene sets.Availability and implementationCode for analyses and figures is available at https://github.com/yexilein/pyroc.Supplementary informationSupplementary data are available at Bioinformatics online.

Les informations bibliographiques du livre des résumés sont à venir après la fin du congrès.; International audience; Feed efficiency is a research priority to support a sustainable meat production, but it is also recognized as a complex trait that integrates multiple biological pathways orchestrated in and by various tissues. This study aims to determine the relationships between biological entities underlying the interindividual variation of feed efficiency in growing pigs. The feed conversion ratio (FCR) was calculated from a total of 47 Large White pigs from a divergent selection for residual feed intake and fed high starch or high-fat high fiber diets during 58 days. We considered transcriptomics (60 k porcine microarray) and metabolomics (1H-NMR analysis and target gas chromatography) datasets obtained in the whole blood of pigs at the end of the trial. We studied whether connecting these two levels of biological organization will help for a better understanding of the key biological processes driving the phenotypic difference in feed efficiency. We identified 33 weighted gene co-expression networks (WGCNA) in the whole blood (13 to 2,491 unique genes). The eigengenes of six modules were correlated to FCR (P < 0.05). Most of these modules were enriched in genes participating in immune and defense-related processes. Profiles in circulating metabolites and fatty acids were summarized by weighed linear combinations from principal component analyses, and correlated to the eigengenes of the co-expressed genes modules. A significant association was found between a module of coexpressed genes participating to T cell receptor signaling and cell development process and related to FCR, and the circulating concentrations of omega-3 fatty acids. Finally, we provided a new method to integrate those experimental data and the public knowledge from metabolic pathway networks databases, to show where and how these molecules at different levels of biological organization participate and interact with each other. Databases of biological pathways in BioPAX format, represented by graphs, were queried to identify participants of interest in interactions. Altogether, these different approaches will help to depict biological systems and understand complex traits such as feed efficiency.

International audience; Alternative splicing of repeats in proteins provides a mechanism for rewiring and fine-tuning protein interaction networks. In this work, we developed a robust and versatile method, ASPRING, to identify alternatively spliced protein repeats from gene annotations. ASPRING leverages evolutionary meaningful alternative splicing-aware hierarchical graphs to provide maps between protein repeats sequences and 3D structures. We re-think the definition of repeats by explicitly accounting for transcript diversity across several genes/species. Using a stringent sequence-based similarity criterion, we detected over 5,000 evolutionary conserved repeats by screening virtually all human protein-coding genes and their orthologs across a dozen species. Through a joint analysis of their sequences and structures, we extracted specificity-determining sequence signatures and assessed their implication in experimentally resolved and modelled protein interactions. Our findings demonstrate the widespread alternative usage of protein repeats in modulating protein interactions and open avenues for targeting repeat-mediated interactions.

This paper addresses the exploration-exploitation dilemma inherent in decision-making, focusing on multi-armed bandit problems. The problems involve an agent deciding whether to exploit current knowledge for immediate gains or explore new avenues for potential long-term rewards. We here introduce a novel algorithm, approximate information maximization (AIM), which employs an analytical approximation of the entropy gradient to choose which arm to pull at each point in time. AIM matches the performance of Infomax and Thompson sampling while also offering enhanced computational speed, determinism, and tractability. Empirical evaluation of AIM indicates its compliance with the Lai & Robbins asymptotic bound and demonstrates its robustness for a range of priors. Its expression is tunable, which allows for specific optimization in various settings.

International audience; Abstract Motivation Boolean networks are simple but efficient mathematical formalism for modelling complex biological systems. However, having only two levels of activation is sometimes not enough to fully capture the dynamics of real-world biological systems. Hence, the need for multi-valued networks (MVNs), a generalization of Boolean networks. Despite the importance of MVNs for modelling biological systems, only limited progress has been made on developing theories, analysis methods, and tools that can support them. In particular, the recent use of trap spaces in Boolean networks made a great impact on the field of systems biology, but there has been no similar concept defined and studied for MVNs to date. Results In this work, we generalize the concept of trap spaces in Boolean networks to that in MVNs. We then develop the theory and the analysis methods for trap spaces in MVNs. In particular, we implement all proposed methods in a Python package called trapmvn. Not only showing the applicability of our approach via a realistic case study, we also evaluate the time efficiency of the method on a large collection of real-world models. The experimental results confirm the time efficiency, which we believe enables more accurate analysis on larger and more complex multi-valued models. Availability and implementation Source code and data are freely available at https://github.com/giang-trinh/trap-mvn.

National audience; Ce chapitre décrit deux applications des méthodes formelles pour Kappa, un langage de réécriture de graphes à sites utilisé en modélisation de systèmes biologiques. La première est une analyse statique pour détecter des motifs qui ne pourront jamais apparaître en cours d'exécution. La seconde réduit les modèles pour obtenir une représentation concise de leurs comportements sous forme de systèmes d'équations différentielles ordinaires.

International audience; The human gut microbiome composition is generally in a stable dynamic equilibrium, but it can deteriorate into dysbiotic states detrimental to host health. To disentangle the inherent complexity and capture the ecological spectrum of microbiome variability, we used 5,230 gut metagenomes to characterize signatures of bacteria commonly co-occurring, termed enterosignatures (ESs). We find five generalizable ESs dominated by either Bacteroides, Firmicutes, Prevotella, Bifidobacterium, or Escherichia. This model confirms key ecological characteristics known from previous enterotype concepts, while enabling the detection of gradual shifts in community structures.Temporal analysis implies that the Bacteroides-associated ES is “core” in the resilience of westernized gut microbiomes, while combinations with other ESs often complement the functional spectrum. The model reliably detects atypical gut microbiomes correlated with adverse host health conditions and/or the presence of pathobionts. ESs provide an interpretable and generic model that enables an intuitive characterization of gut microbiome composition in health and disease.

International audience; One of the main challenges in cancer management relates to the discovery of reliable biomarkers, which could guide decision-making and predict treatment outcome. In particular, the rise and democratization of high-throughput molecular profiling technologies bolstered the discovery of "biomarker signatures" that could maximize the prediction performance. Such an approach was largely employed from diverse OMICs data (i.e., genomics, transcriptomics, proteomics, metabolomics) but not from epitranscriptomics, which encompasses more than 100 biochemical modifications driving the post-transcriptional fate of RNA: stability, splicing, storage, and translation. We and others have studied chemical marks in isolation and associated them with cancer evolution, adaptation, as well as the response to conventional therapy. In this study, we have designed a unique pipeline combining multiplex analysis of the epitranscriptomic landscape by high-performance liquid chromatography coupled to tandem mass spectrometry with statistical multivariate analysis and machine learning approaches in order to identify biomarker signatures that could guide precision medicine and improve disease diagnosis. We applied this approach to analyze a cohort of adult diffuse glioma patients and demonstrate the existence of an "epitranscriptomics-based signature" that permits glioma grades to be discriminated and predicted with unmet accuracy. This study demonstrates that epitranscriptomics (co)evolves along cancer progression and opens new prospects in the field of omics molecular profiling and personalized medicine.

We analyzed changes in the long-term vegetation cover and in fire activity over the past 5000 years in the Ecuadorian páramo using a sediment core from Papallacta (Ecuador). The chronology is constrained by three tephra layers and 32 AMS 14C ages, and 168 samples yielded a high-resolution record of environmental changes. We estimated the upslope wind convectivity as the ratio between pollen transported from the Andean cloud forest and Poaceae pollen to distinguish changes in atmospheric moisture from changes in soil moisture. The record showed that the two sources of moisture, either from year-round adiabatic cloud dripping linked to SASM activity or to ENSO variability at decadal-scale, influenced vegetation-cover changes. Between 5000 and 2450 cal yr BP, both soil moisture and biomass burning were higher than after 2450 cal yr BP. The shift between the two states matches the zonal increase in summer insolation that drove the ITCZ to its southernmost position. Our results underline resilience to volcanic activity, the importance of the upslope convective dripping with the lowest convective index observed at ~4500 cal yr BP, the anomalous last century with the highest convective activity and the driest soil conditions recorded in the last 5000 years, the recent increase in fire activity and the link between soil moisture and the position of the ITCZ.

Cheese organoleptic properties result from complex metabolic processes occurring in microbial communities. A deeper understanding of such mechanisms makes it possible to improve both industrial production processes and end-product quality through the design of microbial consortia. In this work, we caracterise the metabolism of a three-species community consisting ofLactococcus lactis,Lactobacillus plantarumandPropionibacterium freudenreichiiduring a seven-week cheese production process. Using genome-scale metabolic models and omics data integration, we modeled and calibrated individual dynamics using monoculture experiments, and coupled these models to capture the metabolism of the community. This digital twin accurately predicted the dynamics of the community, enlightening the contribution of each microbial species to organoleptic compound production. Further metabolic exploration raised additional possible interactions between the bacterial species. This work provides a methodological framework for the prediction of community-wide metabolism and highlights the added-value of dynamic metabolic modeling for the comprehension of fermented food processes.

The real-world implementation of federated learning is complex and requires research and development actions at the crossroad between different domains ranging from data science, to software programming, networking, and security. While today several FL libraries are proposed to data scientists and users, most of these frameworks are not designed to find seamless application in medical use-cases, due to the specific challenges and requirements of working with medical data and hospital infrastructures. Moreover, governance, design principles, and security assumptions of these frameworks are generally not clearly illustrated, thus preventing the adoption in sensitive applications. Motivated by the current technological landscape of FL in healthcare, in this document we present Fed-BioMed: a research and development initiative aiming at translating federated learning (FL) into real-world medical research applications. We describe our design space, targeted users, domain constraints, and how these factors affect our current and future software architecture.

Background: Single-cell studies have demonstrated the presence of significant cell-to-cell heterogeneity in gene expression. Whether such heterogeneity is only a bystander or has a functional role in the cell differentiation process is still hotly debated. Methods: In this study, we quantified and followed single-cell transcriptional uncertainty – a measure of gene transcriptional stochasticity in single cells – in 10 cell differentiation systems of varying cell lineage progressions, from single to multi-branching trajectories, using the stochastic two-state gene transcription model. Results: By visualizing the transcriptional uncertainty as a landscape over a two-dimensional representation of the single-cell gene expression data, we observed universal features in the cell differentiation trajectories that include: (i) a peak in single-cell uncertainty during transition states, and in systems with bifurcating differentiation trajectories, each branching point represents a state of high transcriptional uncertainty; (ii) a positive correlation of transcriptional uncertainty with transcriptional burst size and frequency; (iii) an increase in RNA velocity preceding the increase in the cell transcriptional uncertainty. Conclusions: Our findings suggest a possible universal mechanism during the cell differentiation process, in which stem cells engage stochastic exploratory dynamics of gene expression at the start of the cell differentiation by increasing gene transcriptional bursts, and disengage such dynamics once cells have decided on a particular terminal cell identity. Notably, the peak of single-cell transcriptional uncertainty signifies the decision-making point in the cell differentiation process.

International audience; Age, sex and APOE-ε4 genotype have been identified as the strongest predictors of the risk of developing Alzheimer's Disease (AD). This work models the pathological progression of regional brain hypometabolism, using mixed-effect models with latent time variable and longitudinal FDG-PET data. Statistical comparisons then disentangle the effects of sex and APOE-ε4 genotype on the onset age and pace of progression of hympometabolism in each brain region, while correcting for education level. They provide a brain map of the regions with earlier and/or faster alterations of the metabolism. We show that females are associated with faster hypometabolism in the caudate nuclei, the thalamus and right temporal and medial-occipital lobes, while APOE-ε4 is associated with earlier hypometabolism in the limbic system (hippocampus, parahippocampus and amygdala) and temporal lobe.

Triatominae are haematophagous bugs that can transmit Trypanosoma cruzi, the causative agent of Chagas disease, endemic to Latin America. In the wild, species belonging to the genus Rhodnius mainly infest palm trees, feeding on a wide range of vertebrate hosts that find refuge in this habitat. However, some Rhodnius species are domiciliary, i.e. they have left their natural habitat (sylvatic environment) and have come into close contact with human populations by colonising their dwellings, and can then transmit T. cruzi to humans. The evolutionary history of species belonging to Rhodnius genus was assessed using a phylogenomic strategy based on whole mitogenome sequences and a set of 51 nuclear genes, characterized for 36 samples representing 16 species out of the 24 described in the genus. This approach allowed us to determine the evolutionary relationship between the 3 major groups of species (pallescens, pictipes and prolixus), to integrate species from Psammolestes genus inside the genus one, and to describe multiple introgression events between species. A comparative genomic approach was conducted between species living in different habitats. The genome size was estimated using flow cytometry and analysis of high throughput sequencing data. Genomes from 11 Rhodnius species (including 4 domiciliary species) were assembled and used to build a calibrated phylogeny including sequences from published genomes and fossil records. The divergence times inferred from this phylogeny allowed us to study genome dynamics. It revealed that the rate of gains and losses per gene per million years is 7.5 times faster in domiciliary species compared to sylvatic ones. This study has made it possible to propose candidate genes including chemosensory receptors involved in the domiciliation of species of the genus Rhodnius. Chemosensory genes belong to multigene families with rapid evolutionary dynamics that cause changes in sensory sensitivity. Their study is especially relevant when searching for genes involved in the recognition of a new environment. The chemosensory receptors were annotated (from 49 to 135 ORs and from 21 to 32 GRs depending on the species studied) and analyzed to detect evidence of selection pressure: 4 candidate genes (OR1, OR38, OR102 and ORco) were identified that could be involved in the domiciliation process. A better understanding of the mechanisms of adaptation of these bugs to their environment could be a key factor in vector control efforts.; Les Triatominae sont des punaises hématophages vectrices de Trypanosoma cruzi, agent responsable de la maladie de Chagas, endémique en Amérique Latine. En milieu naturel, les punaises hématophages du genre Rhodnius infestent principalement les palmiers et se nourrissent sur une large gamme d'hôtes vertébrés qui trouvent refuge dans cet habitat. Cependant, certaines espèces de Rhodnius sont domiciliaires, c'est-à-dire qu'elles ont quitté leur habitat naturel (milieu sylvatique) et sont venues en contact étroit avec les populations humaines en colonisant leurs habitations. Les liens de parenté entre les différentes espèces du genre Rhodnius ont été retracés par une approche phylogénomique reposant sur l'étude des mitogénomes et d'un set de 51 gènes nucléaires et incluant 16 espèces sur les 24 espèces décrites via l'étude de 36 échantillons. Cette approche phylogénomique a permis de statuer sur les affinités évolutives des 3 groupes d'espèces décrits (pallescens, pictipes et prolixus), de classer les espèces du genre Psammolestes au sein du genre Rhodnius et de montrer des événements d'introgression multiples entre espèces. Cette étude phylogénétique a été complétée par une approche de génomique comparative entre espèces adaptées à différents habitats. La taille des génomes a été estimée par deux méthodes, la cytométrie de flux et l'analyse des données moléculaires à haut débit. Pour étudier la dynamique des génomes une phylogénie calibrée à l'aide de données fossiles a été réalisée. Onze génomes de Rhodnius (dont 4 domiciliaires) ont été assemblés puis annotés. L'estimation de la vitesse de gains et pertes de gènes par million d'années montre que ces événements se produisent 7.5 fois plus rapidement chez les espèces domiciliaires en comparaison des espèces sylvatiques. Cette étude a permis de proposer des gènes candidats dont des récepteurs chimiosensoriels impliqués dans la domiciliation des espèces du genre Rhodnius. Les gènes chimiosensoriels appartiennent à des familles multigéniques dont la dynamique évolutive est rapide et à l'origine de changements dans la sensibilité sensorielle. Ils sont donc des objets d'étude de choix dans la recherche de gènes impliqués dans la reconnaissance d'un nouvel environnement. L'annotation spécifique de ces récepteurs chimiosensoriels (de 49 à 135 ORs, de 21 à 32 GRs, selon les espèces) suivie de la recherche de gènes sous sélection a permis de mettre en évidence 4 gènes candidats en lien avec la domiciliation (OR1, OR38, OR102 et ORco). Une meilleure compréhension des mécanismes d'adaptation des punaises hématophages à leur environnement pourrait être un facteur clé dans l'effort de lutte antivectorielle.

There is an ongoing explosion of scientific datasets being generated, brought on by recent technological advances in many areas of the natural sciences. As a result, the life sciences have become increasingly computational in nature, and bioinformatics has taken on a central role in research studies. However, basic computational skills, data analysis, and stewardship are still rarely taught in life science educational programs, resulting in a skills gap in many of the researchers tasked with analysing these big datasets. In order to address this skills gap and empower researchers to perform their own data analyses, the Galaxy Training Network (GTN) has previously developed the Galaxy Training Platform (https://training.galaxyproject.org), an open access, community-driven framework for the collection of FAIR (Findable, Accessible, Interoperable, Reusable) training materials for data analysis utilizing the user-friendly Galaxy framework as its primary data analysis platform. Since its inception, this training platform has thrived, with the number of tutorials and contributors growing rapidly, and the range of topics extending beyond life sciences to include topics such as climatology, cheminformatics, and machine learning. While initially aimed at supporting researchers directly, the GTN framework has proven to be an invaluable resource for educators as well. We have focused our efforts in recent years on adding increased support for this growing community of instructors. New features have been added to facilitate the use of the materials in a classroom setting, simplifying the contribution flow for new materials, and have added a set of train-the-trainer lessons. Here, we present the latest developments in the GTN project, aimed at facilitating the use of the Galaxy Training materials by educators, and its usage in different learning environments.

The rise of single-cell data highlights the need for a nondeterministic view of gene expression, while offering new opportunities regarding gene regulatory network inference. We recently introduced two strategies that specifically exploit time-course data, where single-cell profiling is performed after a stimulus: HARISSA, a mechanistic network model with a highly efficient simulation procedure, and CARDAMOM, a scalable inference method seen as model calibration. Here, we combine the two approaches and show that the same model driven by transcriptional bursting can be used simultaneously as an inference tool, to reconstruct biologically relevant networks, and as a simulation tool, to generate realistic transcriptional profiles emerging from gene interactions. We verify that CARDAMOM quantitatively reconstructs causal links when the data is simulated from HARISSA, and demonstrate its performance on experimental data collected onin vitrodifferentiating mouse embryonic stem cells. Overall, this integrated strategy largely overcomes the limitations of disconnected inference and simulation.Author summaryGene regulatory network (GRN) inference is an old problem, to which single-cell data has recently offered new challenges and breakthrough potential. Many GRN inference methods based on single-cell transcriptomic data have been developed over the last few years, while GRN simulation tools have also been proposed for generating synthetic datasets with realistic features. However, except for benchmarking purposes, these two fields remain largely disconnected. In this work, building on a combination of two methods we recently described, we show that a particular GRN model can be used simultaneously as an inference tool, to reconstruct a biologically relevant network from time-course single-cell gene expression data, and as a simulation tool, to generate realistic transcriptional profiles in a non-trivial way through gene interactions. This integrated strategy demonstrates the benefits of using the same executable model for both simulation and inference.

Complex cellular functions are usually encoded by a set of genes in one or a few organized genetic loci in microbial genomes. Macromolecular System Finder (MacSyFinder) is a program that uses these properties to model and then annotate cellular functions in microbial genomes. This is done by integrating the identification of each individual gene at the level of the molecular system. We hereby present a major release of MacSyFinder (version 2) coded in Python 3. The code was improved and rationalized to facilitate future maintainability. Several new features were added to allow more flexible modelling of the systems. We introduce a more intuitive and comprehensive search engine to identify all the best candidate systems and sub-optimal ones that respect the models' constraints. We also introduce the novel macsydata companion tool that enables the easy installation and broad distribution of the models developed for MacSyFinder (macsy-models) from GitHub repositories. Finally, we have updated and improved MacSyFinder popular models: TXSScan to identify protein secretion systems, TFFscan to identify type IV filaments, CONJscan to identify conjugative systems, and CasFinder to identify CRISPR associated proteins. MacSyFinder and the updated models are available at: https://github.com/gem-pasteur/macsyfinder and https://github.com/macsy-models.

International audience; In recent years, significant technical advances have been made to better observe the structure and the organization of chromatin. Many Hi-C data are now available in public databases, at different accuracy and resolution and for various species. Generally, the final results of Hi-C data analyses are summarized with the representation of “contact maps”. Using a color scale, these maps (heat-maps) show the frequency of contacts observed between different portions of a genome, both at short-rangeand long-range distances.Even if they are widely used, our observation is that the biological interpretation of contact maps is not trivial. It requires training and significant experience to be useful to validate or invalidate hypotheses concerning the overall organization of the studied genome. In this context, alternative tools have been developed. Some of them produce 3D models ofcontact networks, as a substitute for visualization of Hi-C outputs. By 3D models, we mean representations in a 3D space of a genome, in which euclidean distances are derived from the contact frequency matrices. The information contained in the Hi-C data is thus directly translated into distances in a 3D space. Although care must be taken in interpreting these models (they are not “pictures” of the interior of a cell), we observed they are very helpful to highlight specific structural patterns in chromatin organization, originally hidden in contact maps.Developing new strategies to visualize and integrate large scale multi-omics data is the major objective of Thibault Poinsignon’s thesis (ANR MinOmics). These led us to start building 3D models of yeasts andfilamentous fungi genomes, from Hi-C raw sequencing data. We present here our evaluations of the biological significance and the interest to use these models for projecting heterogeneous epigenomics data and hence better support their biological interpretations.

International audience; Aim: The aim of this article is to characterize in detail the γδ T lymphocytes from an adult patient with primary cutaneous T-cell lymphoma of γδ subtype (γδ CTCL). Methods: Here this article reports trajectory mapping on high-resolution differentiation trajectories of γδ T lymphocytes digitally extracted from a scRNAseq dataset. Results: In the patch-to-plaque progression of CTCL, the TCRVγnon9 subset of γδ T cells differentiated from naive T cells (Tn) and central memory T cells (Tcm) to abundant effector memory T cells (Tem) while other cutaneous γδ T and CD8 T cells remained unchanged. Conclusions: This transcriptomic switch underlies the emergence of a CTCL-like progression of the TCRVγnon9 γδ T subtype and suggests new routes for treating these diseases.

The role and biological significance of gene-environment interactions in human traits and diseases remain poorly understood. To address these questions, the CHARGE Gene-Lifestyle Interactions Working Group conducted series of genome-wide interaction studies (GWIS) involving up to 610,475 individuals across four ancestries for three lipids and four blood pressure traits, while accounting for interaction effects with drinking and smoking exposures. Here we used GWIS summary statistics from these studies to decipher potential differences in genetic associations and G×E interactions across phenotype-exposure-ancestry combinations, and to derive insights on the potential mechanistic underlying G×E through in-silico functional analyses. Our analyses show first that interaction effects likely contribute to the commonly reported ancestry-specific genetic effect in complex traits, and second, that some phenotype-exposures pairs are more likely to benefit from a greater detection power when accounting for interactions. It also highlighted modest correlation between marginal and interaction effects, providing material for future methodological development and biological discussions. We also estimated contributions to phenotypic variance, including in particular the genetic heritability conditional on the exposure, and heritability partitioned across a range of functional annotations and cell types. In these analyses, we found multiple instances of potential heterogeneity of functional partitions between exposed and unexposed individuals, providing new evidence for likely exposure-specific genetic pathways. Finally, along this work, we identified potential biases in methods used to jointly meta-analyze genetic and interaction effects. We performed simulations to characterize these limitations and to provide the community with guidelines for future G×E studies.

Most species are extinct, those that are not are often unknown. Sequenced and sampled species are often a minority of known ones. Past evolutionary events involving horizontal gene flow, such as horizontal gene transfer, hybridization, introgression, and admixture, are therefore likely to involve “ghosts,” that is extinct, unknown, or unsampled lineages. The existence of these ghost lineages is widely acknowledged, but their possible impact on the detection of gene flow and on the identification of the species involved is largely overlooked. It is generally considered as a possible source of error that, with reasonable approximation, can be ignored. We explore the possible influence of absent species on an evolutionary study by quantifying the effect of ghost lineages on introgression as detected by the popular D-statistic method. We show from simulated data that under certain frequently encountered conditions, the donors and recipients of horizontal gene flow can be wrongly identified if ghost lineages are not taken into account. In particular, having a distant outgroup, which is usually recommended, leads to an increase in the error probability and to false interpretations in most cases. We conclude that introgression from ghost lineages should be systematically considered as an alternative possible, even probable, scenario. [ABBA–BABA; D-statistic; gene flow; ghost lineage; introgression; simulation.]

Motivation 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 are typically achieved through unsupervised clustering, which crucially relies on a similarity metric. Results 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 high-dimensional data represented as probability distributions. 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 13 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. OT is found to improve cell–cell similarity inference and cell clustering in all simulated and real scRNA-seq data, as well as in scATAC-seq and single-cell DNA methylation data. Availability and implementation All our analyses are reproducible through the OT-scOmics Jupyter notebook available at https://github.com/ComputationalSystemsBiology/OT-scOmics. Supplementary information Supplementary data are available at Bioinformatics online.

International audience; Differentiation can be modeled at the single cell level as a stochastic process resulting from the dynamical functioning of an underlying Gene Regulatory Network (GRN), driving stem or progenitor cells to one or many differentiated cell types. Metastability seems inherent to differentiation process as a consequence of the limited number of cell types. Moreover, mRNA is known to be generally produced by bursts, which can give rise to highly variable non-Gaussian behavior, making the estimation of a GRN from transcriptional profiles challenging. In this article, we present CARDAMOM (Cell type Analysis from scRna-seq Data achieved from a Mixture MOdel), a new algorithm for inferring a GRN from timestamped scRNA-seq data, which crucially exploits these notions of metastability and transcriptional bursting. We show that such inference can be seen as the successive resolution of as many regression problem as timepoints, after a preliminary clustering of the whole set of cells with regards to their associated bursts frequency. We demonstrate the ability of CARDAMOM to infer a reliable GRN from in silico expression datasets, with good computational speed. To the best of our knowledge, this is the first description of a method which uses the concept of metastability for performing GRN inference.

Genome-wide chromosome conformation capture (Hi-C) has revealed the organization of chromatin into topologically associating domains (TADs) and loops, which are thought to help regulate genome functions. TADs and loops are understood as the result of DNA extrusion mediated by the cohesin complex. However, despite recent efforts, direct visualization and quantification of this process in single cells remains an open challenge. Here, we use polymer simulations and dedicated analysis methods to explore if, and under which conditions, DNA loop extrusion can be detected and quantitatively characterized by imaging pairs of fluorescently labeled loci located near loop or TAD anchors in fixed or living cells. We find that under realistic conditions, extrusion can be detected and the frequency of loop formation can be quantified from fixed cell images alone, while the lifetime of loops and the speed of extrusion can be estimated from dynamic live-cell data. Our delineation of appropriate imaging conditions and the proposed analytical methods lay the groundwork for a systematic quantitative characterization of loop extrusion in fixed or living cells.

Démonstration; International audience; Recent advances in high-throughput technologies have revolutionized the analysis in the field of the plant sciences. However, there is an urgent need to effectively integrate and assimilate complementary information to understand the biological system in its entirety. We have developed AgroLD, a knowledge graph that exploits Semantic Web technologies to integrate data of interest for the plant science community e.g., rice, wheat, arabidopsis and in this way facilitate the formulation and validation of new scientific hypotheses. AgroLD contains around 900M triples created by annotating and integrating more than 100 datasets coming from 15 data sources. Our objective is to offer a domain specific knowledge platform to answer complex biological and plant sciences questions related to the implication of genes in, for instance, plant disease resistance or adaptative responses to climate change. In this demo, we present some results which currently focused on genomics, genetics and trait associations.

Numerous models have been developed to account for the complex properties of the random walks of biomolecules. However, when analysing experimental data, conditions are rarely met to ensure model identification. The dynamics may simultaneously be influenced by spatial and temporal heterogeneities of the environment, out-of-equilibrium fluxes and conformal changes of the tracked molecules. Recorded trajectories are often too short to reliably discern such multi-scale dynamics, which precludes unambiguous assessment of the type of random walk and its parameters. Furthermore, the motion of biomolecules may not be well described by a single, canonical random walk model. Here, we develop a two-step statistical testing scheme for comparing biomolecule dynamics observed in different experimental conditions without having to identify or make strong prior assumptions about the model generating the recorded random walks. We first train a graph neural network to perform simulation-based inference and thus learn a rich summary statistics vector describing individual trajectories. We then compare trajectories obtained in different biological conditions using a non-parametric maximum mean discrepancy (MMD) statistical test on their so-obtained summary statistics. This procedure allows us to characterise sets of random walks regardless of their generating models, without resorting to model-specific physical quantities or estimators. We first validate the relevance of our approach on numerically simulated trajectories. This demonstrates both the statistical power of the MMD test and the descriptive power of the learnt summary statistics compared to estimates of physical quantities. We then illustrate the ability of our framework to detect changes in α-synuclein dynamics at synapses in cultured cortical neurons, in response to membrane depolarisation, and show that detected differences are largely driven by increased protein mobility in the depolarised state, in agreement with previous findings. The method provides a means of interpreting the differences it detects in terms of single trajectory characteristics. Finally, we emphasise the interest of performing various comparisons to probe the heterogeneity of experimentally acquired datasets at different levels of granularity (e.g., biological replicates, fields of view, and organelles).

Inferring chemical reaction networks (CRN) from concentration time series is a challenge encouragedby the growing availability of quantitative temporal data at the cellular level. This motivates thedesign of algorithms to infer the preponderant reactions between the molecular species observed ina given biochemical process, and build CRN structure and kinetics models. Existing ODE-basedinference methods such as SINDy resort to least square regression combined with sparsity-enforcingpenalization, such as Lasso. However, we observe that these methods fail to learn sparse modelswhen the input time series are only available in wild type conditions, i.e. without the possibility toplay with combinations of zeroes in the initial conditions. We present a CRN inference algorithmwhich enforces sparsity by inferring reactions in a sequential fashion within a search tree of boundeddepth, ranking the inferred reaction candidates according to the variance of their kinetics on theirsupporting transitions, and re-optimizing the kinetic parameters of the CRN candidates on the wholetrace in a final pass. We show that Reactmine succeeds both on simulation data by retrievinghidden CRNs where SINDy fails, and on two real datasets, one of fluorescence videomicroscopyof cell cycle and circadian clock markers, the other one of biomedical measurements of systemiccircadian biomarkers possibly acting on clock gene expression in peripheral organs, by inferringpreponderant regulations in agreement with previous model-based analyses. The code is available athttps://gitlab.inria.fr/julmarti/crninf/ together with introductory notebooks.

International audience; In the last decades, logical or discrete models have emerged as a successful paradigm for capturing and predicting the behaviors of systems of molecular interactions. Intuitively, they consist in sampling the abundance of each kind of biochemical entity within finite sets of intervals and deriving transitions accordingly. On one hand, formallyproven sound derivation from more precise descriptions (such as from reaction networks) may include many fictitious behaviors. On the other hand, direct modeling usually favors dominant interactions with no guarantee on the behaviors that are neglected. In this paper, we formalize a sound coarse-graining approach for stochastic reaction networks. Its originality relies on two main ingredients. Firstly, we abstract values by intervals that overlap in order to introduce a minimal effort for the system to go back to the previous interval, hence limiting fictitious oscillations in the coarse-grained models. Secondly, we compute for pairs of transitions (in the coarse-grained model) bounds on the probabilities on which one will occur first. We illustrate our ideas on two case studies and demonstrate how techniques from Abstract Interpretation can be used to design more precise discretization methods, while providing a framework to further investigate the underlying structure of logical and discrete models.

Non-cyanobacteria diazotrophs (NCDs) were shown to dominate in surface waters shifting the long-held paradigm of cyanobacteria dominance and raising fundamental questions on how these putative heterotrophic bacteria thrive in sunlit oceans. Here, we report an unprecedented finding in the widely used model diatom Phaeodactylum tricornutum (Pt) of NCDs sustaining diatom cells in the absence of bioavailable nitrogen. We identified PtNCDs using metagenomics sequencing and detected nitrogenase gene in silico and/or by PCR. We demonstrated nitrogen fixation in PtNCDs and their close genetic affiliation with NCDs from the environment. We showed the wide occurrence of this type of symbiosis with the isolation of NCDs from other microalgae, their identification in the environment, and predicted their associations with photosynthetic microalgae. Overall, this study provides evidence for a previously overlooked symbiosis using a multidisciplinary model-based approach, which will help understand the different players driving global marine nitrogen fixation.

International audience; The plant hormone auxin is certainly the most studied developmental regulator in plants. The many functions of auxin during development, from the embryo to the root and shoot construction, are mediated by an ever-growing collection of molecular regulators, with an overwhelming degree of both ubiquity and complexity that we are still far from fully understanding and that biological experiments alone cannot grasp. In this review, we discuss how bioinformatics and computational modeling approaches have helped in recent years to explore this complexity and to push the frontiers of our understanding of auxin biology. We focus on how analysis of massive amounts of genomic data and construction of computational models to simulate auxin-regulated processes at different scales have complemented wet experiments to increase the understanding of how auxin acts in the nucleus to regulate transcription and how auxin movement between cells regulates development at the tissular scale. Expected final online publication date for the Annual Review of Plant Biology, Volume 74 is May 2023. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

We present Svetlana (SuperVised sEgmenTation cLAssifier for NapAri), an open-source Napari plugin dedicated to the manual or automatic classification of segmentation results. A few recent software have made it possible to automatically segment complex 2D and 3D objects such as cells in biology with unrivaled performance. However, the subsequent analysis of the results is oftentimes inaccessible to non-specialists. The Svetlana plugin aims at going one step further, by allowing end-users to label the segmented objects and to pick, train and run arbitrary neural network classifiers. The resulting network can then be used for the quantitative analysis of biophysical phenoma. We showcase its performance through challenging problems in 2D and3D. Comparisons with random forest classifiers, which are the only easily available alternative to date, show significant advantages for the proposed approach.

Richter syndrome (RS) is the transformation of chronic lymphocytic leukemia (CLL) into aggressive lymphoma, most commonly diffuse large B-cell lymphoma (DLBCL). We characterize 58 primary human RS samples by genome-wide DNA methylation and whole-transcriptome profiling. Our comprehensive approach determines RS DNA methylation profile and unravels a CLL epigenetic imprint, allowing CLL-RS clonal relationship assessment without the need of the initial CLL tumor DNA. DNA methylation- and transcriptomic-based classifiers were developed, and testing on landmark DLBCL datasets identifies a poor-prognosis, activated B-cell-like DLBCL subset in 111/1772 samples. The classification robustly identifies phenotypes very similar to RS with a specific genomic profile, accounting for 4.3-8.3% of de novo DLBCLs. In this work, RS multi-omics characterization determines oncogenic mechanisms, establishes a surrogate marker for CLL-RS clonal relationship, and provides a clinically relevant classifier for a subset of primary “RS-type DLBCL” with unfavorable prognosis., publishedVersion

Occupancy models have been extended to account for either multiple spatial scales or species interactions in a dynamic setting. However, as interacting species (e.g., predators and prey) often operate at different spatial scales, including nested spatial structure might be especially relevant to models of interacting species. Here we bridge these two model frameworks by developing a multi-scale, two-species occupancy model. The model is dynamic, i.e. it estimates initial occupancy, colonization and extinction probabilities—including probabilities conditional to the other species’ presence. With a simulation study, we demonstrate that the model is able to estimate most parameters without marked bias under low, medium and high average occupancy probabilities, as well as low, medium and high detection probabilities, with only a small bias for some parameters in low-detection scenarios. We further evaluate the model’s ability to deal with sparse field data by applying it to a multi-scale camera trapping dataset on a mustelid-rodent predator–prey system. Most parameters are estimated with low uncertainty (i.e. narrow posterior distributions). More broadly, our model framework creates opportunities to explicitly account for the spatial structure found in many spatially nested study designs, and to study interacting species that have contrasting movement ranges with camera traps.Supplementary materials accompanying this paper appear online.

Subcellular RNA localization is a critical mechanism for the spatial control of gene expression. Its mechanism and precise functional role is not yet very well understood. Single Molecule Fluorescence in Situ Hybridization (smFISH) images allow for the detection of individual RNA molecules with subcellular accuracy. In return, smFISH requires robust methods to quantify and classify RNA spatial distribution. Here, we present PointFISH, a novel computational approach for the recognition of RNA localization patterns. PointFISH is an attention-based network for computing continuous vector representations of RNA point clouds. Trained on simulations only, it can directly process extracted coordinates from experimental smFISH images. The resulting embedding allows scalable and flexible spatial transcriptomics analysis and matches performance of hand-crafted pipelines.

16 pages, 4 figures, 2 tables, supplementary information https://doi.org/10.1186/s40168-023-01523-z.-- Availability of data and materials: The BBMO microbial sequence abundances (ASV tables), taxonomic classifications, environmental data including nutrients, networks, and R-Markdowns for the data analysis including commands to run eLSA and EnDED (environmentally driven-edge-detection and computing Jaccard index) are publicly available: https://github.com/InaMariaDeutschmann/TemporalNetworkBBMO. DNA sequences are publicly available at the European Nucleotide Archive (https://www.ebi.ac.uk/ena; accession numbers PRJEB23788 for 18S rRNA genes & PRJEB38773 for 16S rRNA genes), Background: Microbial interactions are fundamental for Earth’s ecosystem functioning and biogeochemical cycling. Nevertheless, they are challenging to identify and remain barely known. Omics-based censuses are helpful in predicting microbial interactions through the statistical inference of single (static) association networks. Yet, microbial interactions are dynamic and we have limited knowledge of how they change over time. Here, we investigate the dynamics of microbial associations in a 10-year marine time series in the Mediterranean Sea using an approach inferring a time-resolved (temporal) network from a single static network. Results: A single static network including microbial eukaryotes and bacteria was built using metabarcoding data derived from 120 monthly samples. For the decade, we aimed to identify persistent, seasonal, and temporary microbial associations by determining a temporal network that captures the interactome of each individual sample. We found that the temporal network appears to follow an annual cycle, collapsing, and reassembling when transiting between colder and warmer waters. We observed higher association repeatability in colder than in warmer months. Only 16 associations could be validated using observations reported in literature, underlining our knowledge gap in marine microbial ecological interactions. Conclusions: Our results indicate that marine microbial associations follow recurrent temporal dynamics in temperate zones, which need to be accounted for to better understand the functioning of the ocean microbiome. The constructed marine temporal network may serve as a resource for testing season-specific microbial interaction hypotheses. The applied approach can be transferred to microbiome studies in other ecosystems, Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This project and IMD received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement no. 675752 (ESR2, https://www.singek.eu) to RL. RL was supported by a Ramón y Cajal fellowship (RYC-2013–12554, MINECO, Spain). This work was also supported by the projects INTERACTOMICS (CTM2015-69936-P, MINECO, Spain), MicroEcoSystems (240904, RCN, Norway), and MINIME (PID2019-105775RB-I00, AEI, Spain) to RL. FL was supported by the Spanish National Program FPI 2016 (BES-2016–076317, MICINN, Spain). SC was supported by the CNRS MITI through the interdisciplinary program Modélisation du Vivant (GOBITMAP grant). DE and SC were supported by the H2020 project AtlantECO (award number 862923). A range of projects from the EU and the Spanish Ministry of Science funded the data collection and ancillary analyses at the BBMO. We acknowledge funding of the Spanish government through the “Severo Ochoa Centre of Excellence” accreditation (CEX2019-000,928-S)

International audience

International audience; CRISPR-based genome editing systems have been successfully and effectively used in many organisms. However, only a few studies have reported the comparison between CRISPR/Cas9 and CRISPR/Cpf1 systems in the whole-genome applications. Although many web-based toolkits are available, there is still a shortage of comprehensive, user-friendly, and plant-specific CRISPR databases and desktop software. In this study, we identified and analyzed the similarities and differences between CRISPR/Cas9 and CRISPR/Cpf1 systems by considering the abundance of proto-spacer adjacent motif (PAM) sites, effects of GC content, optimal proto-spacer length, potential universality within the plant kingdom, PAM-rich region (PARR) inhibiting ratio, and effects of G-quadruplex (G-Q) structures. Using this information, we built a comprehensive CRISPR database (including 138 plant genome data sources, www.grapeworld.cn/pc/index.html), which provides search tools for the identification of CRISPR editing sites in both CRISPR/Cas9 and CRISPR/Cpf1 systems. We also developed a desktop software on the basis of Perl/TK tool, which facilitates and improves the detection and analysis of CRISPR editing sites at the whole-genome level on Linux and/or Windows platform. Therefore, this study provides helpful data and software for easy selection and application of CRISPR-based genome editing systems in plants.

International audience; The COVID-19 pandemic has led to a resurgence of the debate on whether host-parasite interactions should evolve towards avirulence. In this review, we first show that SARS-CoV-2 virulence is evolving, before explaining why some expect the mortality caused by the epidemic to converge towards that of human seasonal alphacoronaviruses. Leaning on existing theory, we then include viral evolution into the picture and discuss hypotheses explaining why the virulence has increased since the beginning of the pandemic. Finally, we mention some potential scenarios for the future.

The collection of animal position data via GPS tracking devices has increased in quality and usage in recent years. Animal position and movement, although measured discretely, follows the same principles of kinematic motion, and as such, the process is inherently continuous and differentiable. I demonstrate the functionality and visual elegance of smoothing spline models. I discuss the challenges and benefits of implementing such an approach, and I provide an analysis of movement and social interaction of seven jaguars inhabiting the Taiamã Ecological Station, Pantanal, Brazil, a region with the highest known density of jaguars. In the analysis, I derive measures for pairwise distance, cooccurrence, and spatiotemporal association between jaguars, borrowing ideas from density estimation and information theory. These measures are feasible as a result of spline model estimation, and they provide a critical tool for a deeper investigation of cooccurrence duration, frequency, and localized spatio‐temporal relationships between animals. In this work, I characterize a variety of interactive relationships between pairs of jaguars, and I particularly emphasize the relationships in movement of two male–female and two male–male jaguar pairs exhibiting highly associative relationships., I propose the use of smoothing spline models to explore animal movement and social interaction. I propose temporal‐state density estimation and localized mutual information measures to characterize behavior and strength of association between jaguars on a refined and unified time grid. I provide further characterization of time‐dependent male–male and female–male relationships in this region, an important step in determining conservation policy for an endangered neotropical predator.

The human ADAMTS-TSL multidomain proteins are involved in numerous pathologies. Encoded by 26 paralogous genes, their domain combination is not sufficient to characterize their functional differences. We propose in this thesis a new approach to identify functional regions of the sequences. For this purpose, we use sequences from 9 eukaryotic species to identify conserved sequence modules specific to certain subgroups of homologous sequences. The evolutionary analysis of the identified modules is obtained by performing a joint phylogenetic reconstruction of genes, species and modules. Furthermore, to validate the functional interest of the identified modules, we associate phenotypes (PPI) to this evolutionary history. This has led to the identification of concomitant acquisitions of "modules/phenotypes", predicting the functionality of these modules. Applying this approach to human ADAMTS-TSL proteins has allowed us to identify new, finer, non-contiguous functional regions that can describe their specificities.; Les protéines multidomaines ADAMTS-TSL humaines sont impliquées dans de nombreuses pathologies. Codées par 26 gènes paralogues, leur combinatoire en domaines ne suffit pas à caractériser leurs différences fonctionnelles. Nous proposons dans cette thèse une nouvelle approche d'identification des régions fonctionnelles des séquences. Pour cela nous utilisons des séquences de 9 espèces eucaryotes afin d'identifier des modules de séquences conservées propres à certains sous-groupes de séquences homologues. L'analyse évolutive des modules identifiés est obtenue en effectuant une reconstruction phylogénétique conjointe des gènes, des espèces et des modules. Par ailleurs, pour valider l'intérêt fonctionnel des modules identifiés, nous associons des phénotypes (PPI) à cette histoire évolutive. Ce qui a abouti à identifier des acquisitions concomitantes de « modules/phénotypes », prédisant la fonctionnalité de ces modules. Appliquer cette approche aux protéines ADAMTS-TSL humaines nous a permis d'identifier de nouvelles régions fonctionnelles, plus fines, non contiguës et à même d'en décrire les spécificités.

Interactions between proteins are one of the foundations of the development of life and their identification and understanding are still major elements of fundamental and applied research. In this context, the focus is on post-translational modifications of proteins that can alter their efficiency and lifetime. In addition, specific interactions between proteins can now be studied at the atomic level thanks to the development of experimental methods for solving the structures of protein complexes. However, these methods still do not always provide the expected results and their cost, whether financial or in terms of time, may prevent the understanding of certain phenomena, particularly during the emergence of a health crisis such as COVID-19. This is why, in parallel, computational methods such as molecular docking or molecular dynamics have been developed. This thesis is situated in these two contexts: firstly, the prediction of O-GlcNAcylation sites, a post-translational modification, catalyzed by a single enzyme called OGT, which has been extensively studied and implicated in different diseases such as cancer, Alzheimer's disease and type 2 diabetes. Secondly, in the context of COVID-19, interactions between human and viral proteins were highlighted through a world-wide study, in which the CAPRI protein docking experiment proposed several of these interactions to expert modelers of protein complexes in order to better understand the mechanisms of COVID-19.The prediction of O-GlcNAcylation sites is not a new research field, as some tools for this type of prediction already exist. We have created a new data set, in order to compare and differentiate these. As the different algorithms consistently showed too many false positives, we developed an improvement based on a larger dataset but also on structural characteristics. However, the results still show too much heterogeneity to allow a safe prediction. Additional results support the theory that chaperone proteins are required for the enzyme to recognise its substrate. In order to better understand the mechanisms of this modification, the interaction between beta-catenin and OGT was specifically studied. This interaction has been shown to be involved in colorectal cancer and is therefore of particular interest.To establish the veracity of the proposed models for the interactions between the human and SARS-CoV-2 proteins, a method based on the consensus of all the models produced was developed. Initial test results showed this method to be effective. We therefore tested its predictive capacity on a new and larger dataset provided by CAPRI. Once again, the developed method showed good results. It was then compared with pre-existing scoring algorithms on a similar benchmark and demonstrated improved results. The method also showed that the interaction models between viral and human proteins are not as reliable as desired.; Les interactions entre les protéines sont l'une des bases du développement de la vie. Leur identification et compréhension sont toujours des éléments majeurs de la recherche fondamentale et appliquée. Dans cette optique, on s'intéresse aux modifications post-traductionnelles des protéines qui ont la capacité d'altérer leur efficacité et leur durée de vie. Les interactions spécifiques entre protéines sont désormais étudiées au niveau atomique grâce au développement des méthodes expérimentales pour résoudre des structures de complexes protéiques. Cependant, ces méthodes ne permettent toujours pas d'obtenir les résultats escomptés et leur coût, que ce soit financier ou en termes de temps, peut empêcher la compréhension de certains phénomènes, notamment lors d'émergence de crise sanitaire comme le COVID-19. C'est pourquoi, en parallèle, des méthodes informatiques telles que l'amarrage moléculaire ou la dynamique moléculaire ont été développées. Cette thèse se situe dans ces deux contextes: dans un premier temps, la prédiction de sites de O-GlcNAcylation, une modification post-traductionnelle, catalysée par une seule enzyme appelée OGT, très étudiée qui est impliquée dans différentes maladies telles que le cancer, la maladie d'Alzheimer et le diabète de type 2. Dans un second temps, et ceci dans le contexte du COVID-19, des interactions entre les protéines humaines et virales ont été mises en avant mais avec la montée rapide de cas d'infection et les méthodes expérimentales étant trop longues, une expérimentation mondiale appelée CAPRI a proposé plusieurs des ces interactions aux modélisateurs du monde entier.La prédiction de sites de O-GlcNAcylation n'est pas une recherche récente car des outils proposent déjà cette possibilité. Afin de les comparer, une base de données a été créée pour les différencier. Comme les différents logiciels montraient un trop grand nombre de faux positifs, une amélioration basée sur cette plus grande base de données mais aussi sur des caractéristiques structurelles a été proposée. Malgré cela, les résultats montrent une trop grande hétérogénéité pour permettre une prédiction sûre. Des résultats supplémentaires appuient la théorie du besoin de protéines auxiliaires pour permettre à l'enzyme la reconnaissance de son substrat. Afin de mieux comprendre les mécanismes de cette modification, l'interaction entre la beta-caténine et l'OGT a été étudiée spécifiquement. En effet, cette interaction a été montrée comme étant impliquée dans le cancer colorectal et révèle donc un intérêt particulier.Pour établir la véracité des modèles proposés pour les interactions entre les protéines du SARS-CoV-2 et de l'humain, une méthode basée sur le consensus de tous les modèles produits a été développée. Au vu des premiers résultats, cette méthode semblait performante. C'est pourquoi sa capacité de prédiction a été testée sur une nouvelle grande base de données, fournie par CAPRI. Une fois encore, la méthode développée a montré de bons résultats. Elle a ensuite été comparée aux logiciels de scoring actuels et montre ici de meilleurs résultats. Hélas, cette méthode montre que les modèles d'interaction entre les protéines virales et humaines ne sont pas aussi fiables que souhaités.

The emergence of new sequencing technologies, as long-read sequencing, has paved the way for the production of reference genomes at an affordable cost. Bioinformatics tools, as well as laboratory techniques, are constantly evolving and it is important to test and implement state-of-the-art technologies.This thesis presents, through different publications, the methodological developments implemented in the framework of reference genome assemblies. The long reads obtained using the Oxford Nanopore sequencing technology allowed us to improve the contiguity of our assemblies and to accurately reconstruct regions that were resistant to older sequencing technologies, as regions containing repeated sequences or clusters of duplicated genes. The optical map technology proposed by Bionano Genomics and the sequencing of Hi-C libraries, allowed us to produce assemblies at the chromosome scale and to access the structure of complex regions such as centromeres.I have been able to demonstrate the importance of these enhanced assemblies with different plant genomes, for which they provide a valuable resource for future research on evolutionary history, adaptations to new living conditions or resistance genes.These developments will be adapted to meet the needs of future large-scale projects aimed at producing reference genome and preserving the genetic information of biodiversity.; L'avènement des nouvelles technologies de séquençage puis du séquençage longues lectures a ouvert la voie vers la production de génomes de référence à un coût abordable. Les outils bioinformatiques, tout comme les techniques de laboratoire, ne cessent d'évoluer et il est important de tester et mettre en place toutes ces technologies de pointe.Ce mémoire de thèse présente, au travers de différentes publications, les développements méthodologiques mis en place dans le cadre d'assemblages de génomes de référence. Les longues lectures obtenues grâce au séquençage avec la technologie Oxford Nanopore nous ont permis d'améliorer la continuité de nos assemblages et de reconstituer plus fidèlement des régions souvent invisibles avec des méthodes traditionnelles, notamment les régions contenant des séquences répétées ou des clusters de gènes dupliqués. La technologie de carte optique proposée par Bionano Genomics, et le séquençage de banques Hi-C nous ont permis d'amener nos assemblages à l'échelle des chromosomes et d'accéder à la structure de régions dites complexes comme les centromères.J'ai pu mettre en évidence l'importance de ces nouveaux assemblages pour différents génomes de plantes, pour lesquels ils constituent une ressource très précieuse pour de futures recherches sur l'histoire évolutive des espèces, sur les adaptations à de nouvelles conditions de vie ou sur les gènes de résistance aux pathogènes.Ces développements vont être adaptés pour répondre aux besoins de nouveaux projets de production de génomes de référence à grande échelle et dont le but est de préserver l'information génétique de la biodiversité.

We introduce TiPS, an R package to generate trajectories and phylogenetic trees associated with a compartmental model. Trajectories are simulated using Gillespie’s exact or approximate stochastic simulation algorithm, or a newly-proposed mixed version of the two. Phylogenetic trees are simulated from a trajectory under a backwards-in-time approach (i.e. coalescent). TiPS is based on the Rcpp package, allowing to combine the flexibility of R for model definition and the speed of C++ for simulations execution. The model is defined in R with a set of reactions, which allow capturing heterogeneity in life cycles or any sort of population structure. TiPS converts the model into C++ code and compiles it into a simulator that is interfaced in R via a function. Furthermore, the package allows defining time periods in which the model’s parameters can take different values. This package is particularly well suited for population genetics and phylodynamics studies that need to generate a large number of phylogenies used for population dynamics studies. This package is available on the CRAN at https://cran.r-project.org/package=TiPS.

International audience; Presentation is an overview of the projects developed at GDEC Institute on wheat genomics and bioinformatics. Bread wheat is one of the most important crops and improving our knowledge about its genome and the diversity of cultivated varieties and wild relatives is needed to face the challenge of a more sustainable agriculture. Our bioinformatics research activities concern wheat genome sequencing, gene and transposon annotation, and analysis of structural variations. The particularity of wheat genomics is the complexity of its genome which is 15 Gb, hexaploid, and shaped by millions of copies of transposons. We develop strategies and methodologies dedicated to tackle the complexity of this genome in order to get a clear view of its composition, a better understanding of transposon dynamics, and the impact of polyploidy. Wheat scientific community has now reached the pangenomics area, with multiple high-quality genome sequences available. An overview of our bioinformatic developments and recent results about wheat pangenomics will be presented.

The analysis of large quantities of biomedical data makes it possible to obtain better results and helps researchers and clinicians to establish good diagnoses and make the right decisions. National and international collaborations that explore precision medicine are based on analyzes of data from diverse and numerous populations. Current biomedical analyzes are subject to problems in terms of data protection. In this multidisciplinary thesis, we contribute to the resolution of these problems through the development of distributed biomedical data analyses as well as their deployment in IT infrastructures and their integration into real medical tools. As part ofa kidney transplant application, we provide a new distributed contextualization solution that helps clinicians assess patients’ kidney problems. For the same application, we also propose a new distributed version of Factor Analysis of Mixed Data (FAMD) for dimension reduction. In the context of HLA database, we propose a new distributed and secure model for frequency estimation applied to HLA databases. We also propose a new algorithm for the distributed estimation of HLA haplotype frequency using the EM algorithm. All our contributions have been deployed on a real geo-distributed cloud infrastructure and evaluated on real or realistic data. Experiments have shown (very) good results in terms of performance compared to centralized solutions.; L’analyse de larges quantités de données biomédicales permet d’obtenir une meilleure précision des résultats et aide les chercheurs et les cliniciens à établir de bonnes diagnostiques et prendre les bonnes décisions. Les collaborations nationales et internationales qui explorent la médecine de précision se fondent sur des analyses sur des données provenant de populations variées et nombreuses. Les analyses biomédicales actuelles sont sujets à des problèmes en terme de protection de données.Dans cette thèse multidisciplinaire, nous contribuons à la résolution de ces problèmes par le développement d’analyses de données biomédicales distribuées ainsi que par leur déploiement dans des infrastructures informatiques et leur intégration dans des outils médicaux réels. Dans le cadre d’une application de transplantation rénale, nous proposons une nouvelle solution de contextualisation distribuée qui aide les cliniciens à évaluer les problèmes rénaux de patients. Pour la même application nous proposons également une nouvelle version distribuée de l’analyse factorielle des données mixtes (FAMD) pour la réduction des dimensions. Dans le cadre des données HLA, nous proposons un nouveau modèle distribué et sécurisé pour l’estimation de fréquences appliquées sur des bases de données HLA. Nous proposons également un nouveau algorithme pour l’estimation distribuée de la fréquence d’haplotypes HLA en utilisant l’algorithme EM. Toutes nos contributions ont été déployées sur une infrastructure réelle de nuage géodistribués et évaluées sur des données réelles ou réalistes. Les expérimentations ont montré de (très) bons résultats en terme de performances comparés aux solutions centralisées.