2023Activity reportProject-TeamCAPSID

3.1.1 Context

In this axis, the CAPSID team develops methods related to knowledge discovery from databases (KDD, 36). The diversity of biological databases and resources is such today that it is more and more difficult to consider each database independently from the others  62. A limited subset of these resources is devoted to the 3D structure of biological objects (proteins, nucleic acids, glycanes...). Structural information is also contained in databases classifying protein domains as building blocks of proteins that can be reused in different proteins sharing the same function (Pfam, CATH and InterPro are well-known examples of such databases)  56, 67, 26. There are millions of proteins across all living species but only tens of thousands of domains that are combined in proteins. Thus, complex tasks such as predicting protein function or interactions can be simplified when envisaged at the domain level.

Due to the great diversity of databases, Knowledge Graphs (KGs) are more and more used to represent and integrate biological information. There is no single definition of KGs as these graphs cover a large variety of domains and data representation contexts (for instance the GAFAM companies advertize various KG uses). The main feature that differentiates KGs from classical graphs is the fact that both nodes (or entities) and edges (or relations) in the graph are heterogeneous and belong to various types described in the KG schema (metagraph). The field of biological knowledge discovery in KG is expanding rapidly  52. Most biological KGs today are developed for drug repurposing tasks (e.g. HetioNet  41 or DRKG). Clinicians are also very interested in network science carried out on rich knowledge graphs as a mean to interpret biomarker studies. However, there is still a need for curated, reliable biological KGs and for efficient knowledge discovery methods in KGs.

3.1.2 Knowledge discovery from protein structural databases

Concerning protein structural databases, we aim to explore novel classification paradigms exploiting existing resources about protein folds and domains  21, 22, 56, 67. In particular it will be interesting to use Kpax, our structural alignment tool  63, to define domain-domain similarity matrices. A non-trivial issue with clustering is how to handle similarity rather than mathematical distance matrices, and how to find the optimal number of clusters, especially when using a non-Euclidean similarity measure. We will adapt the algorithms and the calculation of quality indices to the Kpax similarity measure. More fundamentally, it will be necessary to integrate this classification step in the more general KDD process leading from data to knowledge.

For example, protein domain classification is relevant for studying domain-domain interactions (DDI). Our previous work on Knowledge-Based Docking (KBDOCK, 38, 40) will be updated and extended using newly published DDIs. Methods for inferring new DDIs from existing protein-protein interactions (PPIs) will be developed. Efforts should be made for validating such inferred DDIs so that they can be used to enrich DDI classification and predict new PPIs.

3.1.3 Function Annotation in Large Protein Graphs

Knowledge of the functional properties of proteins can shed considerable light on how they might interact. However, huge numbers of protein sequences in public databases such as UniProt/TrEMBL lack any functional annotation, and the functional annotation of such sequences is a highly challenging problem. We are developing graph-based and machine learning techniques to annotate automatically the available unannotated sequences with functional properties such as Enzyme Commission (EC) numbers and Gene Ontology (GO) terms (note that these terms are organised hierarchically allowing generalization/specialization reasoning). The idea is to transfer annotations from expert-reviewed sequences present in the UniProt/SwissProt database (about 560 thousands entries) to unreviewed sequences present in the UniProt/TrEMBL database (about 80% of 180 millions entries). For this, we have to learn from the UniProt/SwissProt database how to compute the similarity of proteins sharing identical or similar functional annotations. Various similarity measures can be tested using cross-validation approches in the UniProt/SwissProt database. For instance, we can use primary sequence or domain signature similarities. More complex similarities can be computed with graph-embedding techniques.

3.1.4 Knowledge discovery algorithms in large biological knowledge graphs

KGs are particularly useful and appropriate in biology, to represent and integrate the complex contents of biological databases  60. We intend to design algorithms for leveraging information embedded in biological KGs (also known as complex networks). In biology, KGs mostly represent PPIs, integrated with various properties attached to proteins, such as pathways, drug binding or relation with diseases. Setting up similarity measures for proteins in a knowledge graph is a difficult challenge. Our objective is to extract useful knowledge from such graphs in order to better understand and highlight the role of multi-component assemblies in various types of cell or organisms. Ultimately, knowledge graphs can be used to model and simulate the functioning of such molecular machinery in the context of the living cell, under physiological or pathological conditions.

3.2.1 Context

This axis deals with 3D protein structure and interactions. In fact, the long-lasting problem of predicting a 3D structure from a protein sequence has been solved in 2021 by the AlphaFold2 (DeepMind)  46 or RosettaFold methods  24. This success, revealed in the CASP14 (Critical Assessment of Stucture Prediction) challenge, was possible not only thanks to AI methods but also because the amount of experimental 3D structures has reached a sufficient size in the Protein Data Bank (PDB). For the same type of reasons, the rigid docking problem (in which the bodies to dock are rigid) seems to be on the way to being solved as well  28, 35. However, research is still required to address the problem of docking disordered proteins or flexible nucleic acids that will fold as they bind to proteins. This is the direction taken by the team since the arrival of Isaure Chauvot de Beauchêne, the inventor of a fragment-based approach for RNA docking onto proteins.

Modeling protein - and even more RNA - flexibility accurately during docking is very computationally expensive. This is due to the very large number of internal degrees of freedom in each molecule, associated with twisting motions around covalent bonds. Therefore, it is highly impractical to use detailed force-field or geometric representations in a brute-force docking search. Instead, most docking algorithms use fast heuristic methods to perform an initial rigid-body search in order to locate a relatively small number of candidate binding orientations, and these are then refined using a more expensive interaction potential or force-field model, which might also include flexible refinement using molecular dynamics (MD), for example.

3.2.2 Coarse-Grained Models

Many approaches have been proposed in the literature to take into account protein (and more recently RNA/DNA) flexibility during docking. The most thorough methods rely on expensive atomistic simulations using MD. However, much of a MD trajectory is unlikely to be relevant to a docking encounter unless it is constrained to explore a putative protein-protein/NA interface. Consequently, MD is normally only used to refine a small number of candidate rigid body docking poses. A much faster - but more approximate - method is to use "coarse-grained" (CG) normal mode analysis (NMA) techniques to reduce the number of flexible degrees of freedom to just one or a handful of the most significant vibrational modes  34, 54, 57, 59. In our experience, docking ensembles of NMA conformations do not give much improvement over basic FFT-based soft docking  71, and it is very computationally expensive to use side-chain repacking to refine candidate soft docking poses  39.

In the last few years, CG force-field models have become increasingly popular in the MD community because they allow very large biomolecular systems to be simulated using conventional MD programs  23. Typically, a CG force-field representation replaces the 5-15 atoms in each amino acid with 2-4 “pseudo-atoms” (each pseudo-atom represents few atoms of an amino-acid as a single bead). It then assigns each pseudo-atom a small number of parameters to represent its chemo-physical properties. By directly attacking the quadratic nature of pair-wise energy functions, coarse-graining can speed up MD simulations by up to three orders of magnitude. Nonetheless, such CG models can still produce useful models of very large multi-component assemblies  66. Furthermore, this type of CG model effectively integrates many internal DOFs to build a smoother but still physically realistic energy surface  42. We are currently developing a CG scoring function for RNA-protein docking by fragments assembly.

3.2.3 Assembling Multi-Component Complexes and Integrative Structure Modelling

We also want to develop related approaches for integrative structure modeling using cryo-electron microscopy (cryo-EM). Thanks to recent developments in cryo-EM instruments and technologies, it is now feasible to capture low resolution images of very large macromolecular machines. However, while such developments offer the intriguing prospect of being able to trap biological systems in unprecedented levels of detail, there will also come with an increasing need to analyse, annotate, and interpret the enormous volumes of data that will soon flow from the latest instruments. In particular, a new challenge that is emerging is how to fit previously solved high resolution protein structures into low resolution cryo-EM density maps. However, the problem here is that large molecular machines will have multiple sub-components, some of which will be unknown, and many of which will fit each part of the map almost equally well. Thus, the general problem of building high resolution 3D models from cryo-EM data is like building a complex 3D jigsaw puzzle in which several pieces may be unknown or missing, and none of which will fit perfectly. We wish to proceed firstly by putting more emphasis on the single-body terms in the scoring function  33, and secondly by using fast CG representations and knowledge-based distance restraints to prune large regions of the search space, as initiated with the EROS-DOCK software  6, 65.

3.2.4 Protein-Nucleic Acid Interactions

As well as playing an essential role in the translation of DNA into proteins, RNA molecules carry out many other essential biological functions in cells, often through their interactions with proteins. A critical challenge in modeling such interactions computationally is that the RNA is often highly flexible, especially in single-stranded (ssRNA) regions of its structure. These flexible regions are often very important because it is through their flexibility that the RNA can adjust its 3D conformation in order to bind to a protein surface. However, conventional protein-protein docking algorithms generally assume that the 3D structures to be docked are rigid, and so are not suitable for modeling protein-RNA interactions. There is therefore much interest in developing dedicated protein-RNA docking algorithms which can take RNA flexibility into account. This research topic has been initiated with the recruitement of Isaure Chauvot de Beauchêne in 2016 and is becoming a major activity in the team. A novel flexible docking algorithm is currently under development in the team. It first docks small fragments of ssRNA (typically three nucleotides at a time) onto a protein surface, and then combinatorially reassembles those fragments in order to recover a contiguous ssRNA structure on the protein surface  29, 30.

As the correctness of the initial docking of the fragments settles an upper limit to the correctness of the full model, we are now focusing on improving that step. A key component of our docking tool is the energy function of the protein-fragment interactions that is used both to drive the sampling (positioning of the fragments) by minimization, and to discriminate the correct final positions from decoys (i.e., false positives). We are developing a new approach to create knowledge-based parameters for coarse-grain energy functions from public structural data, in collaboration with Sjoerd de Vries (INSERM). Such approach will be applied first to ssRNA-protein complexes, then to other types of complexes such as protein-peptides.

Another key requirement for this approach is an exhaustive but non-redundant library of possible internal conformations of RNA fragments. Our library is built by clustering hundreds of thousands of experimentally known RNA structures, based on an approximate geometric similarity criteria. We want to develop new algorithms for the clustering of 3D conformations based on internal coordinates and on epsilon-net theory, in order to optimise the representativity and computational cost of the library.

In the future, we will improve the combinatorial algorithm used for reassembling the docked fragments using both experimental constraints and knowledge-based constraints pertaining from the research carried out in Axis 1.

Allosteric pathways in protein-nucleic acid complexes.

Novel methodological approaches based on deep learning (AlphaFold2 and RosettaFold) have started to make remarkable advances in protein structure prediction and design. However, our knowledge regarding their dynamical behaviour and function is still highly limited. One important example is the notion of allostery which refers to processes whereby a binding event at one site of a biological macromolecule affects the binding activity at another distinct functional site, enabling the regulation of the corresponding function. The allosteric behavior of a macromolecular system arises from the properties of the native free-energy landscape of the system, and how this landscape is remodeled by various perturbations, such as ligand binding, protonation, mutations, post-translational modifications, or interactions with other molecules. Therefore, understanding allosteric pathways of communication is of high importance and is not well understood yet. All-atom MD simulations could be used to capture subtle dynamical changes that are associated with allosteric signaling. Moreover, graph theory-based methods were developed to investigate the set of trajectories generated by MD simulations and extract allosteric pathways. Yasaman Karami has previously developed a method called COMMA (COMmunication Mapping) that describes the dynamical architecture of a protein by predicting the network of communications within the system  48 and she is constantly working at improving it.

In 2023, Benjamin Gottis was hired for an M2 internship (6 months, February-July) to improve COMMA for the analysis of large protein complexes. His internship resulted in the improvement of threshold definitions in COMMA and accelerating the calculations. Moreover, Yasaman Karami and Malika Smaïl-Tabbone recruited a PhD student, Victor Pryakhin who obtained a doctoral contract from Université de Lorraine and started in October 2023. His PhD subject is to investigate communication networks in protein-RNA complexes using deep learning approaches. At the same time, Yasaman Karami has obtained computational resources on Jean Zay super computer to perform MD simulations on a set of protein-RNA complexes. The results of this computations will be used in Viktor Pryakhin's doctoral project.

Dynamics of Type IV Pilus.

Type IV pili (T4P) are dynamic filaments at the surface of many bacteria that can rapidly extend and retract and withstand strong forces. T4P are responsible for various cellular processes due to their highly dynamic behavior and are crucial for bacterial virulence in many human pathogens, including Enterohemorrhagic Escherichia coli (EHEC), Pseudomonas aeruginosa (PaK), Neisseria meningitidis (Nm), Neisseria gonorrhoeae (Ng), and Myxococcus xanthus (Mx). They are assembled by complex protein machinery localized in the bacterial envelope, formed by the repeat of major pilin subunits organised in a helical manner. The structure of these T4Ps have been determined by combining NMR with a cryo-EM density map of the pilus, resulting in an atomistic model of the T4Ps filament. Despite the recent progress in cryo-EM and integrative structural biology, the high flexibility of this family of fibers often limits the resolution of the structure. Modeling is therefore a necessary step in structure determination. To unravel the structural basis of different T4Ps dynamics, we performed extensive and large-scale classical and steered MD simulations of five different T4Ps, EHEC, PaK, Nm, Ng and Mx. The results highlighted key regions for the function of those filaments. Such simulations require an important number of computational resources, which were provided thanks to the Swiss National Supercomputing Centre (CSCS) and GENCI computing center. This project is one of the main activities of Yasaman Karami, in collaboration with Michael Nilges (Institut Pasteur, Paris) and Edward Egelman (University of Virginia, USA).

6.1.1 CROMAST

• Name:
  Cross-Mapper of domain Structural instances
• Keywords:
  Protein domain, 3D structure, Classification, Databases, Workflow
• Scientific Description:
  see the scientific paper 10.1093/bioadv/vbad081
• Functional Description:
  CroMaSt (Cross Mapper of domain Structural instances) is an automated iterative workflow to clarify the assignment of protein domains to a given domain type of interest, based on their 3D structure and by cross-mapping of domain structural instances between domain databases. CroMaSt (for Cross-Mapper of domain Structural instances) will classify all structural instances of a given domain type into 4 different categories (Core, True, Domain-like, and Failed)
• Release Contributions:
  Operational version for publication
• URL:
  https://­workflowhub.­eu/­workflows/­390
• Publication:
  hal-04210856
• Contact:
  Hrishikesh Dhondge

6.1.2 HIPPO

• Name:
  HIstogram-based Pseudo-POtential
• Keywords:
  Structural Biology, Computational biology
• Functional Description:
  Pipeline to create scoring potentials from docking decoys
• Contact:
  Isaure Chauvot De Beauchêne

6.1.3 RRMpip

• Name:
  RRM_modeling_pipeline
• Keywords:
  Structural Biology, Computational biology
• Functional Description:
  pipeline to create 3D models of RRM protein domains
• Contact:
  Isaure Chauvot De Beauchêne

6.1.4 ssRNATTRACT

• Keywords:
  Computational biology, Structural Biology
• Functional Description:
  3D modelling of protein-ssRNA complexes by fragment assembly
• URL:
  https://­github.­com/­isaureCdB/­ssRNATTRACT
• Contact:
  Isaure Chauvot De Beauchêne

7.1.1 Knowledge graph mining with embedding-based methods

In the context of Md Kamrul Islam's PhD project, we addressed the problem of link prediction in large knowledge graphs (KGs) using KG embedding methods. These methods aim to learn low dimensional vector representations of entities and relations in a KG. Such representations (in a latent space) facilitate link prediction tasks along with other downstream tasks. In this context, it is important to achieve both an efficient KG embedding and explainable predictions. During learning of efficient embeddings, sampling negative triples is an important step as KGs only come with observed positive triples. We previoulsy proposed an efficient simple negative sampling (SNS) method based on the assumption that the entities which are closer to the corrupted entity in the embedding space are able to provide high-quality negative triples  43. As for explainability, we also report in the same paper a new rule mining method which exploits the learned embeddings  43.

We then extended this work to propose an integrated drug repurposing, evaluation and explanation pipeline for COVID-19 disease 11. The workflow starts with collecting and cleaning a COVID-19 centric drug repurposing knowledge graph (DRKG). Then, high-quality and compact ensemble embeddings are learned using three embedding methods. The embeddings are then used to train a deep neural network based model to predict the probability of unobserved triples connecting drugs with 27 COVID-19 proteins (as drug targets). The top-100 predictions are evaluated based on (i) cross-matching with in-trial drugs for COVID-19 and (ii) molecular evaluation based on compound and protein structures. Beside these evaluations, we learn high quality rules from DRKG and provide possible explanations of predictions. This study demonstrates how complementary embedding methods can be used to generate high-quality ensemble embeddings of a KG and how to use embeddings for the drug repurposing task. To the best of our knowledge, it is the first attempt to combine virtual screening methods with KG embedding methods in predicting and evaluating repurposable drugs for COVID-19. Besides the retrieval of many in-trial drugs, both methods show a converging result that the Fosinopril compound could be a new potential nsp13 inhibitor. Experimental validation of Fosinopril compound to treat COVID-19 is a potential perspective of this study. The molecular evaluation results and explanations of the predictions make us confident about the drawn conclusions. Md Kamrul Islam has successfully defended his PhD on December 16, 2022  44.

7.1.2 Biological network modeling

Boolean Networks (BNs) refer to a simple formalism used to study complex biological systems when the prediction of exact reaction times is not of interest. BNs play a key role in understanding the dynamics of the studied systems and in predicting their disruption in case of complex human diseases. The BioModels database is a well-known repository of peer-reviewed models represented in the Systems Biology Markup Language (SBML). Most of these models are quantitative, but in some use cases, qualitative models—such as BNs—are better suited. In the context of Athénais Vaginay's PhD project, we proposed SBML2BN, a pipeline dedicated to the automatic transformation of quantitative SBML models to Boolean networks  69. Our approach takes advantage of several SBML elements (reactions, rules, events) as well as a numerical simulation of the concentration of the species over time to constrain both the structure and the dynamics of the Boolean networks to synthesise. Finding all the BNs complying with given structure and dynamics was formalised as an optimisation problem formulated in the answer-set programming framework.

We ran SBML2BN on more than 200 quantitative SBML models, and we could construct Boolean networks which are compatible with the structure and the dynamics of the SBML models 69. The most recent work relies on abstract simulation of a chemical reaction network (CRN) to avoid the tricky binarization task as we propose to simulate chemical reaction networks with the deterministic semantics abstractly, without any precise knowledge on the initial concentrations. For this, the concentrations of species are abstracted to Booleans stating whether the species is present or absent, and the derivatives of the concentrations are abstracted to signs saying whether the concentration is increasing, decreasing, or unchanged. We use abstract interpretation over the structure of signs for mapping the ODEs of a reaction network to a Boolean network with nondeterministic updates. The abstract state transition graph of such Boolean networks can be computed by finite domain constraint programming over the finite structure of signs. Constraints on the abstraction of the initial concentrations can be added naturally, leading to an abstract simulation algorithm that produces only the part of the abstract state transition graph that is reachable from the abstraction of the initial state. We proved the soundness of our abstract simulation algorithm, and showed its applicability to reaction networks in the SBML format from the BioModels database 14. Athénaïs Vaginay has successfully defended her PhD on July, 7 2023 16.

7.1.3 Machine Learning and Scalable Graph-based Approaches

In the context of a collaboration with researchers from both the University Badji Mokhtar-Annaba (Algeria), the University of Quebec At Montreal (UQAM) and the University of Lille, Sabeur Aridhi proposed a genetic algorithm for random forest 12. The proposed algorithm has three main objectives: (1) strengthening the classification accuracy of individual decision trees as well as that of the forest, (2) making use of diversity measures among the decision trees to improve the generalization of the constructed model, and (3) minimizing the number of trees in the forest and finding an optimal subset of the random forest.

7.2.1 Modeling and design of RNA-RRM complexes

Our recent H2020 ITN project RNAct (2018-2022) aimed at designing new RNA-binding proteins based on the evolutionary conserved protein domain1, called RNA Recognition Motif (RRM). In this context and in the continuity of this project, we develop approaches to create 3D models of ssRNA-RRM complexes, building up on our expertise in RNA-protein modeling by fragments assembly  30. We use the ATTRACT docking software to sample and evaluate possible (low binding energy) positions of RNA fragments on the protein surface, then assemble the geometrically compatible poses with compatible sequences, to build a continuous model for the full RNA sequence.

7.2.2 3D Modeling of proteins and protein complexes

7.2.3 Investigating the dynamical behaviour of protein complexes with MD simulation

7.2.4 Machine learning methods for proteomics, interactomics and protein design

7.2.5 Miscellaneous results on structural studies of host-pathogen interactions

8.1.1 Participation in other International Programs

8.2.1 Visits of international scientists

8.2.2 Visits to international teams

8.3.1 Other european programs/initiatives

Through her implication in the French institute of Bioinformatics (joint coordination of the Open Science and Interoperability taskforce), Marie-Dominique Devignes is a member of the European ELIXIR Interoperability platform.

ANR EPIHLA

Participants: Marie-Dominique Devignes [contact person], Malika Smaïl-Tabbone, Diego Amaya Ramirez, Bernard Maigret.

• Title:
  HLA compatibility in organ transplantation : from antigens to epitopes (EPIHLA)
• Duration:
  October 2022-October 2025
• Coordinator:
  Pr. Jean-Luc Taupin (Inserm U976, Saint-Louis Hospital, Paris)
• Inria contact:
  Marie-Dominique Devignes
• Partner Institutions:
  • Inserm U976 IRSL Saint-Louis Hospital (Paris)
  • LORIA CNRS (Nancy)
  • INSERM U1016 Cochin Institute (Paris)
  • CNRS U144 Institut Curie (Paris)
• Summary:
  The EPIHLA project has two major aims. (1) It aims at correctly representing HLA molecule 3D structure and superimposing predicted conformations in order to identify 3D differences that could constitute epitopes and eplets, targets of donor-specific antibodies. (2) It aims at developing the capacity to isolate and clone anti-HLA antibody genes from patients’ B lymphocytes. The results will provide decisive new information on the understanding of humoral alloreactivity and will make it possible to better anticipate transplant rejection. This project was initially based on the Inria-Inserm PhD project of Diego Amaya Ramirez (2019-2022). This thesis ("HLA genetic system and organ transplantation: understanding the basics of immunogenicity to improve donor - receptor compatibility when assigning grafts to recipients") is not finished yet and still co-supervised by Marie-Dominique Devignes and Pr. Jean-Luc Taupin.

ANR GlycoLink

Participants: Hamed Khakzad [contact person], Yasaman Karami, Isaure Chauvot de Beauchêne.

• Title:
  Exploring Glycosyltransferases complexes involved in glycosaminoglycan-Linker assembly
• Duration:
  October 2023-October 2026
• Coordinator:
  Pr. Sylvie Fournel-Gigleux (UMR 7365 CNRS-Université de Lorraine)
• Inria contact:
  Hamed Khakzad
• Partner Institutions:
  • UMR 7365 CNRS-Université de Lorraine (Nancy)
  • IBS - UMR 5075 (Grenoble)
  • INRIA Center of Université de Lorraine (Nancy)
  • ICOA, UMR 7311 CNRS U-Orléans (Orléans)
• Summary:
  The supramolecular arrangement of glycosyltransferases (GT) and accessory enzymes (kinase, phosphatase, sulfotransferases) governs the repertoire of carbohydrates and their multiple biological functions. The assembly of glycosaminoglycans (GAG), a major class of linear glycopolymers, is initiated by a unique tetrasaccharide linker (GlcAb1-3Galb1-3Galb1-4Xyl) synthesized by the coordinated action of five GT. It serves as a unique primer for the elongation of GAG chains. A recent paradigm puts forward a multimolecular complex called “GAGosome" formed of GT and auxiliary enzymes as a major mechanism to guarantee fidelity and efficiency of GAG synthesis. Indeed, association of the heparan-sulfate (HS) polymerases EXT1/EXT2 and of chondroitin-sulfate (CS) synthases (CSS) sustain GAG elongation. Recent evidence also supports complex formation between GAG-linker enzymes. GlycoLink will explore the formation and organization of multimolecular complexes between GT and partners involved in the assembly of the GAG-linker region, and their functional impact in rare genetic diseases.

LUE-FEDER CITRAM (2017-2023)

The CITRAM project (Conception d'Inhibiteurs de la Transmission de Résistances AntI-Microbiennes), co-funded by Lorraine Université d'Excellence (LUE) and FEDER was extended until June 2023.

Partners other than CAPSID are:

• DynAMic lab (Genome dynamics and microbial adaptation, INRAE-Université de Lorraine UMR 1128), Team of Nathalie Leblond, coordinator ;
• LPCT (Laboratoire de Physique et Chimie Théoriques, CNRS-Université de Lorraine UMR 7019), Team of Chris Chipot.

MolAI4Cryo (2023-2025)

The molAI4Cryo project (Modeling and Artificial Intelligence applied to Cryo-EM 3D structures to fight COVID) has obtained a grant from the Région Grand-Est to equip the IMoPA laboratory with a cutting-edge equipement in cryo-electromicroscopy. The CAPSID team will be involved in proposing algorithms to use cryo-EM results as constraints for modeling protein-RNA complexes.

Partners other than CAPSID are:

• IMoPA (Ingénierie Moléculaire et Physiopathologie Articulaire, CNRS-Université de Lorraine UMR 7365), Team of Xavier Manival, coordinator ;
• DynAMic lab (Genome dynamics and microbial adaptation, INRAE-Université de Lorraine UMR 1128), Team of Nathalie Leblond and Nicolas Soler ;
• LPCT (Laboratoire de Physique et Chimie Théoriques, CNRS-Université de Lorraine UMR 7019), Team of Chris Chipot and François Dehez.

9.1.1 Scientific events: organisation

Yasaman Karami organized in Nancy (12 July 2023) a seminar called NCSB (Nancy Computational Structural Biology) with the help from both Inria and Loria. The event was held successfully with about 50 participants at the Université de Lorraine. We had the pleasure of having Professor Edward Egelman as the keynote speaker. In total we had 7 oral communications by both regional (IMoPA, DynaAMic, LPCT, Inria-Loria) and international (Germany, USA) scientists.

9.1.2 Invited talks

Marie-Dominique Devignes gave an invited talk on "Open Science at the French Institute of Bioinformatics" during the Love Data Week organized at the Université de Lorraine, 13-16 March 2023.

Yasaman Karami gave an invited talk on “Conformational dynamics of Type-IV Pilus” during the Statistical Physics and Low Dimensional systems (SPLDS) Conference organized by the LPCT lab at Pont-a-Mousson, 24-26 May, 2023.

9.1.3 Leadership within the scientific community

Marie-Dominique Devignes is co-leader of the Open Science and Interoperability Task at the French Institute of Bioinformatics (with Alban Gaignard - Institut du Thorax, Nantes, Platform BiRD, and Frederic de Lamotte - INRAE, Department of Biology et Plant Improvement, Montpellier). In this frame, she is member of the ELIXIR European Interoperability platform and she co-authored a paper on automatic FAIR assessment of web resource data 9.

Hamed Khakzad is an active member of the Rosetta Commons scientific community.

9.1.4 Scientific expertise

Sabeur Aridhi reviewed an ANR project in 2023.

9.1.5 Research administration

Malika Smaïl-Tabbone is member of the Scientific Council of the Université de Lorraine. As such, she carries out various scientific expertises within the broad framework of the University of Lorraine.

9.2.1 Teaching

• Malika Smaïl-Tabbone is an associate professor at the Université de Lorraine with a full service. She is co-responsible with Pascal Moyal of the IMSD track ("Ingénierie Mathématique pour la Science des Données") in the Applied Mathematics Master's degree at the Université de Lorraine. She is also a member of the pedagogic team of the CMI BSE ("Cursus Master Ingénieur Biologie-Santé-Environnement").
• Sabeur Aridhi is an assistant professor at the Université de Lorraine with a full service. He is responsible for the major in IAMD ("Ingénierie et Applications des Masses de Données") at TELECOM Nancy.
• Marie-Dominique Devignes teaches every year 10 to 16h in the CMI BSE.
• Diego Amaya Ramirez held an ATER position from January to August 2023.
• Yasaman Karami gave 27 hours on Machine learning with Python as part of the Data Scientist continuing education course at the Institute for Digital Management and Cognition (IDMC), Nancy.
• Hamed Khakzad gave 16 hours on deep learning methods in the frame of the Advanced AI course (3A) at TELECOM Nancy.

9.2.2 Supervision

In 2023, there have been in total 7 PhD students supervised by CAPSID members. Three of them successfully defended their thesis. Three of them have been recruited and started their PhD in October 2023. Hamed Khakzad is also co-supervising with Rebekka Wild a PhD thesis that started in October 2023 at the University of Grenoble.

9.2.3 Juries

Malika Smaïl-Tabbone was reviewer of two PhD theses.

Marie-Dominique Devignes was reviewer of one PhD thesis and examiner of two PhD theses.

(Not counting the participation of team members in the juries of CAPSID PhD students)

9.2.4 Internal or external Inria responsibilities

Yasaman Karami is a member of the Inria Comité de Développement Technologique (CDT). The main task of the committee is to evaluate the application and recruitement of research engineers.

International journals

• 7 articleM.-D.Marie-Dominique Devignes, M.Malika Smaïl-Tabbone, H.Hrishikesh Dhondge, R.Roswitha Dolcemascolo, J.Jose Gavaldá-García, R. A.R. Anahí Higuera-Rodriguez, A.Anna Kravchenko, J.Joel Roca Martínez, N.Niki Messini, A.Anna Pérez-Ràfols, G.Guillermo Pérez Ropero, L.Luca Sperotto, I.Isaure Chauvot de Beauchêne and W.Wim Vranken. Experiences with a training DSW knowledge model for early-stage researchers.Open Research Europe32023, 97HALDOIback to text
• 8 articleH.Hrishikesh Dhondge, I.Isaure Chauvot de Beauchêne and M.-D.Marie-Dominique Devignes. CroMaSt: a workflow for assessing protein domain classification by cross-mapping of structural instances between domain databases and structural alignment.Bioinformatics Advances31January 2023HALDOIback to text
• 9 articleA.Alban Gaignard, T.Thomas Rosnet, F.Frédéric de Lamotte, V.Vincent Lefort and M.-D.Marie-Dominique Devignes. FAIR-Checker: supporting digital resource findability and reuse with Knowledge Graphs and Semantic Web standards.Journal of Biomedical Semantics141July 2023, 7HALDOIback to text
• 10 articleC.Carlos Gueto-Tettay, D.Di Tang, L.Lotta Happonen, M.Moritz Heusel, H.Hamed Khakzad, J.Johan Malmström and L.Lars Malmström. Multienzyme deep learning models improve peptide de novo sequencing by mass spectrometry proteomics.PLoS Computational Biology191January 2023, e1010457HALDOIback to text
• 11 articleM. K.Md Kamrul Islam, D.Diego Amaya-Ramirez, B.Bernard Maigret, M.-D.Marie-Dominique Devignes, S.Sabeur Aridhi and M.Malika Smaïl-Tabbone. Molecular-evaluated and explainable drug repurposing for COVID-19 using ensemble knowledge graph embedding.Scientific Reports131March 2023, 3643HALDOIback to text
• 12 articleN. E.Nour El Islem Karabadji, A.Abdelaziz Amara Korba, A.Ali Assi, H.Hassina Seridi, S.Sabeur Aridhi and W.Wajdi Dhifli. Accuracy and diversity-aware multi-objective approach for random forest construction.Expert Systems with Applications2251April 2023, 120138HALDOIback to text
• 13 articleH.Hamed Khakzad, I.Ilia Igashov, A.Arne Schneuing, C.Casper Goverde, M.Michael Bronstein and B.Bruno Correia. A new age in protein design empowered by deep learning.Cell Systems1411November 2023, 925-939HALDOIback to text

International peer-reviewed conferences

• 14 inproceedingsJ.Joachim Niehren, C.Cédric Lhoussaine and A.Athénaïs Vaginay. Core SBML and its Formal Semantics.CMSB 2023 - 21th International Conference on Formal Methods in Systems BiologyLuxembourg, LuxembourgSeptember 2023HALback to text

Doctoral dissertations and habilitation theses

• 15 thesisH.Hrishikesh Dhondge. Structural characterization of RNA binding to RNA recognition motif (RRM) domains using data integration, 3D modeling and molecular dynamic simulation.Université de LorraineJuly 2023HALback to text
• 16 thesisA.Athénaïs Vaginay. Synthesis of Boolean networks from the structure and dynamics of reaction networks.Université de LorraineJuly 2023HALback to text

Other scientific publications

• 17 inproceedingsD.Diego Amaya-Ramirez, R.Romain Lhotte, C.Cedric Usureau, M.Magali Devriese, M.Malika Smaïl-Tabbone, J.-L.Jean-Luc Taupin and D.Devignes Marie-Dominique. HLA-EpiCheck : A B-cell epitope prediction tool on HLA antigens using molecular dynamics simulation data.ISMB-ECCB 2023 - Intelligent System in Molecular Biology and European Conference on Computational Biology merged eventLyon, FranceJuly 2023HALback to text
• 18 inproceedingsA.Anna Kravchenko, S.Sjoerd Jacob de Vries, M.Malika Smaïl-Tabbone and I.Isaure Chauvot de Beauchene. HIPPO: HIstogram-based Pseudo-POtential for scoring ssRNA-protein fragment-based docking poses.The 31st Annual Intelligent Systems For Molecular Biology (ISMB) and the 22nd Annual European Conference on Computational Biology (ECCB)Lyon, FranceJuly 2023HALback to text