Team: DISCO (DIgital Systems and COmputational cancer)
NEW
New Master’s Program : Digital Health Device Design (Starting Septembre 2024)
https://www.univ-lille.fr/formations/fr-00171115
Co-responsable : Dr. D. Zitouni & Dr W. Dhifli
The study of regulatory networks (signaling pathways, transcriptional and metabolic networks) is emerging as a major approach in understanding and managing cancers. This systemic perspective allows for a better understanding of the role of genetic and epigenetic modifications, as well as the stability and plasticity of the regulatory network in the tumor phenotype, and the sensitivity or resistance to treatments.
Leveraging principles of systems biology and Artificial Intelligence techniques, our computational DISCO team contribute to:
• Bridge the gap between high-throughput biological data production and analytical tools to enhance our understanding of the systemic principles of tumor biology.
• Investigate cell regulatory networks to unravel cancer tumorigenesis, cellular heterogeneity, and therapy resistance.
• Develop and disseminate AI-powered Bio-IT tools that scientists can readily understand, trust, and effectively utilize for biological investigations.
Towards these goals, we design and develop new algorithms and software for building and analyzing regulatory networks from ‘omics’ data, network-based interpretation and integration of biological systems, network causality, and network-based hypothesis generation. Our approaches range from machine learning, evolutionary computation, and knowledge representation to closed-loop bioscience automation.
We collaborate closely with experimental groups and share our tools as free, open-source packages and web applications. e.g., network inference/reconstruction: LICORN/CoRegNet, PreCisIon; network analysis/interrogation: PEPPER, skyclust, LatNET, cRegMap; network intervention/revision: Adalab, CoRegFlux, GREAT.
Our team receives significant external funding from the European Union and the French National Science Foundation (EC Horizon2020, EC FP7 Chist-Era, INERM-ITMO Cancer, INCa-PLBIO, ANR…).
We regularly welcome master’s and doctoral students, postdoctoral researchers, and engineers. If you would like to join us, please send us your application.
> RESEARCHERS
> INGENEERS/TECHNICIANS
> STUDENTS
RESEARCHERS
| Pr Mohamed ELATI, Université de Lille mohamed.elati(@)univ-lille.fr |
| Dr Wajdi DHIFLI, Maître de Conférence, Université de Lille wajdi.dhifli(@)univ-lille.fr |
| Dr Nour KARABADJI, Maître de Conférence, Université de Badji Mokhtar Annaba noud.karabadji(@)univ-lille.fr |
| Dr Djamel ZITOUNI, Maître de Conférence, Université de Lille Research field: Knowledge representation, visual analytics, software development, Health informatics.. djamel.zitouni(@)univ-lille.fr |
| Dr Chloé BERNHARD, Project lead , INCa EN-HOPE SMART4CBT Research field: Project and data sharing, Bioinformatics, systems biology, molecular classification. chloe.bernhard(@)univ-lille.fr |
| Dr Konstantinos GELES, Post-Doctoral fellow, Université de Lille Research field: computational biology, multi-omics, data integration, molecular classification. geles.konstantinos(@)univ-lille.fr |
INGENEERS/TECHNICIANS
| Jules DUSOL, Study Engineer, Université de Lille jules.dusol(@)univ-lille.fr |
| Nadia BENNACEUR, Research Technician, Université de Lille nadia.bennaceur(@)univ-lille.fr |
STUDENTS
| Geoffrey PAWLAK (D3, PhD) geoffrey.pawlak.etu(@)univ-lille.fr |
| Nathan LANERET (D2, PhD)nathan.laneret.etu(@)univ-lille.fr |
| Ines HADJADJI (D2, PhD)ines.hadjadhji.etu(@)univ-lille.fr |
| Liangwei YIN (D1, PhD).etu(@)univ-lille.fr |
Funded Projects
Europe:
2021-2024 : GREENER: Gene and Regulatory Elements Networks Involved in Rice Root Tissue Differentiation. – PRCI ANR/FNR – Coordinator Dr. Christophe Perin (CIRAD, Montpellier), Role : PI Univ. Lille-CANTHER – 800 K€. (own funding 195K€).
2015-2018 : Adaptive « Automated Scientific Laboratory – AdaLab » – CHIST-ERA – Coordinators Prof. Larisa Soldatova (Brunel University), Prof. Ross King (Univ. of Manchester) – 1300 K€. Role : ANR coordinateur, PI UEVE-iSSB (198K€).
2012-15: “A system biology approach to dissect cilia function and its disruption in human genetic disease” – SYSCILIA (EC FP7 – Health Cooperation programme). Coordinator Dr. Ronald Roepman (Univ of Nigmegen) – €1200K. Role: PI UEVE-iSSB (200K€)
National:
2024-2028: Siric Pediatric: En-Hope Smart4cbt. “East North-Hematology Oncology PEdiatric consortium offering a research program of Social sciences, Microenvironment and multiomics Analysis in RadioTherapy resistance For Children Brain Tumors”. 2024-2028. Budget: €3 million. Role Tasks co-leader.
2022-2024 : NETMET: Regulation network of the oncogene addiction triggers by MET receptor in lung cancer, INCa PLBIO. Role PI Univ. Lille-CANTHER (own funding: 58 K€)
2018-2022 : INTEGRIN: Systems biology of integrin inhibition-induced apoptosis for novel glioblastoma treatment, INCa PLBIO, Grant N2017-145. Role PI Univ. Lille-CANTHER (own funding: 115 K€)
2015-2019 : LIONs: Large-scale Integrative approach to unravel the relationships between differentiatiON and tumorigenesiS. ITMO cancer/INSERM 2015 – coordinator: Mohamed Elati 750 K€. Role: coordinator PI Univ Lille-CANTHER (220K€)
2016-2019 : CHASSY: From multi-scale modeling of biological network to engineering metabolic circuits in a biotechnology chassis, Paris-Saclay IDEX, Grant Programme Interdisciplinaire IDI (co-PI), Horizon 2020 No 7208242014-16: « Ingénierie robuste et évolution dirigée de voies métaboliques synthétiques par intégration des approches génomique » coordinator. Dr. François Jean-Marie (INSA Toulouse) – 500€K Rôle: co- PI François Képès (170 K€).
2013-2014 : “Comparaison de Réseaux de régulation par Enumération de PErturbations – CREPE (PEPS CNRS). Coordinator Pr. Etienne Birmelé (Univ. Paris 5) – 20 K€. Rôle: PI UEVE-iSSB (8K€)
2011-2013 : “Search for new therapeutic targets through the Identification of Networks Specifically altered during tumorIGenesis” – INSIGht (INCa). Coordinator Dr. François Radvanyi (Institut Curie) – €464K. Rôle: PI UEVE-iSSB (115 K€)
Team Projects
CANTHER.XAI – eXplainable AI engine to accelerate CANcer systems THERapeutics
We propose to develop CANTHER.XAI (eXplainable AI engine to accelerate CANcer systems THERapeutics) to rationalize cancer systems therapeutics research through an AI closed-loop system (Fig.1). This loop comprises three fundamental components:
- Inference (from experimental cellular models)
- Interrogation (to patients)
- Intervention (feedback to cellular models) with regulatory networks.
By utilizing this loop alongside regulatory networks, we bridge the gap between cellular models and human levels to construct a large-scale manifold of disease states using machine learning embedding techniques. As of today, while large molecular datasets are technically available publicly, this data is not accessible to all and not interoperable. CANTHER-XAI aims to aggregate and help the scientific community exploit all available datasets. This large-scale manifold will provide us with crucial insights into master regulators, matched cellular models, actionable interventions, and patient populations likely to respond positively to treatments.
Our vision is for CANTHER.XAI to synergize with scientists and clinicians (see Fig.2), enabling:
- Conducting system-level research with shared models on comparable data from meta-cohorts at least 10 times larger than current possibilities,
- Maximizing the number of discovered and experimentally verified hypotheses, and
- Acting as a collaborator.
Although primarily designed for fundamental cancer systems biology research, discoveries made with CANTHER.XAI can have significant clinical impacts, such as identifying therapeutic targets, molecular classifications, and facilitating drug repurposing efforts.Specifically, our key objectives include:
- Obj1. Developing a knowledge base, CANTHER.KB, and machine learning/network tools, CANTHER.AI, for applying CANTHER.XAI closed-loop concept in cancer systems biology.
- Obj2. Incorporating explainability into CANTHER.XAI, facilitating seamless interaction between scientists and the AI engine.
- Obj3. Evaluating CANTHER.SaaS (Software as a Service) within collaborative organizations of biologists and clinicians to advance systems and precision oncology
The access to CANTHER-XAI through a web browser (i.e. as SaaS) has been confirmed as an excellent solution both for the users and for developers. We anticipate CANTHER.XAI to feature five main outputs (Fig.2):
- CoRegNet: We will use and extend a conceptual tumor regulatory architecture in which the aberrant activity of master regulators is both necessary and sufficient to maintain tumor cell state. Within a tumor, the nature of the network and its activity vary from cancer cells due to genetic and epigenetic heterogeneity, cell location within a tumor, and the presence of stromal cells. The network approach proposed in this project is original mainly in terms of the incorporation into the model of the factor of cooperativity between co-regulators (Elati et al., Bioinformatics 2007, Nicolle et al., Bioinformatics 2015), rendering it closer to the biological reality than most other approaches.
- CoReglink: The translation of TF co-regulatory networks into graphical models and mapping with signaling (Zerrouk et al., Scientific Reports 2020, Miagoux et al., Journal of Personalized Medicine 2021) and/or metabolic (Trejo et al., BMC Systems Biology 2017, Coutant et al., PNAS 2019) networks should also make the joint simulation of regulatory network models possible.
- CoRegMap: Each cancer type is a heterogeneous disease. Several types of heterogeneity have been described: intertumor heterogeneity revealed by the identification of molecularly diverse subgroups of tumors, and intratumor heterogeneity. We will pursue our work on constructing a large-scale manifold of disease states from latent network-based representations (Picchetti et al., BMC Bioinformatics 2015, Dhifli et al., BMC Bioinformatics 2019), the identification and the characterization of molecular subgroups and their matched cellular models.
- CoRegTreat: We will focus on the identification of therapeutic targets and guide the repositioning of drugs based on the paradigm of reversing transcriptional regulatory networks. Therapeutic compounds will not be assessed by their binding affinity to a particular protein, but by predicting their ability to induce a transcriptional response that is anti-correlated with the regulatory program underlying the pathological state.
- CoRegPan: Comparing regulatory networks of different tissues should help in understanding the specificity that may exist in these tissues. Comparing primary vs metastatic provides insight into how cancer tumors adapt to their metastatic environments. Tumor cells use and modify the networks already present in the normal cells of origin. It is therefore of interest to compare networks in tumors and those in the matched normal cells (Champion et al., Bioinformatics and Comp. Biology, 2021).
PUBLICATIONS
PREPRINTS
Ch. Bernhard, K. Geles, G. Pawlak, W. Dhifli, A. Dispot, M. Kondratova, J. Dusol, N. Entz-Werle, S. Martin, M. Messé, D. Reita, D. Tulasne, I. Van Seuningen, S-A Ciafrè, M. Dontenwill, M. Elati. A co-regulatory influence map of glioblastoma heterogeneity and evolution – From cells in vitro to tumors and back again.
M-J Truong, G. Pawlak, J-P Meneboo, S Sebda, M. Fernandes, M. Figeac, M. Elati and D. Tulasne. Transcriptional program-based deciphering of the MET exon 14 skipping regulatory network.
Moreno-Vega*, M. Zambrano*, J. Puig*, F. Dufour, X. Meng, J. Fontugne, C. Groeneveld, M. Shi, S. Lindskrog, C. Beraud, E. Chapeaublanc, Th. Lebret, A. Eijan, Y. Allory, Ph. Luel, L. Dyrskjot, M. Elati, F. Radvanyi, C. Lodillinsky and I.Bernard-Pierrot, Identification of a pro-tumoral role of p63 in altered-FGFR3 tumors through inference of bladder cancer gene regulatory network and functional validations.
N. Karabadji, A. Korba, A. Assi, H. Seridi, M. Karabadji, Y. Ghamri-Doudane, A. Lakhdari, M. Elati, W. Dhifli. A Genetic and Graph-Guided Feature Learning Strategy for Improving Decision Tree Construction.
I. Hadjadji, N. Karabadji, H. Seridi, N. Manaa, M. Elati, W. Dhifli. PSO and Graph-Guided Approach for Automatic Optimization of Neural Network Architecture.
L. Yin, P. Traversa, M. Elati, Y. Moreno, N. Marek-Trzonkowska, Ch. Battail. A comprehensive analysis of sample-specific network features for predicting immunotherapy response in advanced kidney cancer.
M. Messé, Ch. Bernhard, S. Foppolo, L. Thomas, P. Marchand, Ch. Herold-Mende, A. Idbaih, H. Kessler, N. Etienne-Selloum, Ch. Ochoa, U. Tambar, P. Laquerriere, N. Entz-Werle, S. Martin, M. Elati, D. Reita, and M. Dontenwill. Hypoxia-driven heterogeneous expression of 5 integrin in glioblastoma stem cells is linked to HIF-2α.
B. Caron, N. Jonckheere, J. Biakou, A. Thévenot, G. Averous, M. Elati, A. Nair, E. Martin, M. Vanier, V. Rebours, A. Couvelard, J. Cros, P. Bachellier, C. Domon-Dell, I. Van Seuningen, M. Plateroti, J-M Reimund, M Tavian, J-N Freund, I Duluc. Abnormal ectopic expression of the CDX2 transcription factor in the pancreas is pro-oncogenic with different effects of its DNA-binding and transactivation domains.
2024
Liu J, Wang X, Jiang W, Azoitei A, Eiseler T, Eckstein M, Hartmann A, Stilgenbauer S, Elati M, Hohwieler M, Kleger A, John A, Wezel F, Zengerling F, Bolenz C, Günes C. Impairment of α-tubulin and F-actin interactions of GJB3 induces aneuploidy in urothelial cells and promotes bladder cancer cell invasion. Cell Mol Biol Lett. 2024 Jul 2;29(1):94. (DOI: https://doi.org/10.1186/s11658-024-00609-2 )
2023
N.E.I. Karabadji, A. A. Korba, A. Assi, H. Seridi, S. Aridhi, W. Dhifli. Accuracy and Diversity-Aware Multi-Objective Approach for Random
Forest Construction. Expert Systems With Applications (Elsevier) 225:120138, 2023. (DOI: https://doi.org/10.1016/j.eswa.2023.120138 )
Wang, X., Liu, J., Azoitei, A., Eiseler, T., Meessen, S., Jiang, W., Elati M, … & Günes, C. (2023). Loss of ORP3 induces aneuploidy and
promotes bladder cancer cell invasion through deregulated microtubule and actin dynamics. Cellular and Molecular Life Sciences, 80(10). (DOI: https://doi.org/10.1007/s00018-023-04959-6 )
Fruchart M, El Idrissi F, Lamer A, Belarbi K, Lemdani M, Zitouni D, Guinhouya BC. Identification of early symptoms of endometriosis through the analysis of online social networks: A social media study. Digit Health. 2023;9:20552076231176114. (DOI: https://doi.org/10.1177/20552076231176114 )
Fruchart, M., El Idrissi, F., Lamer, A., Belarbi, K., Lemdani, M., Zitouni, D., & Guinhouya, B. (2023). Une approche patient-centrée pour
l’identification des symptômes précoces de l’endométriose – fouille de texte des réseaux sociaux en ligne. Journal of Obstetrics and Gynaecology (Vol. 45, Issue 5, p. 356). Elsevier BV. (DOI: https://doi.org/10.1016/j.jogc.2023.03.050 )
2022
K. M. Pékpé, D. Zitouni, G. Gasso, W. Dhifli, B.C. Guinhouya. From SIR to SEAIRD: a novel data-driven modeling approach based on the Grey System Theory to predict the dynamics of COVID-19. Applied Intelligence (Springer) 1-10, 2022 (DOI: https://doi.org/10.1007/s10489-021-02379-2 )
Miagoux, Q., Singh, V., Dereck, D. M., Chaudru, V., Elati, M., Teixeira, E., & Niarakis, A. (2021). De la création d’un réseau intégratif
spécifique de la polyarthrite rhumatoïde à l’étude d’un modèle personnalisé relié au traitement des patients. Revue du Rhumatisme, 88, A151. (DOI : https://doi.org/10.1016/j.rhum.2021.10.238 )
El Idrissi, F., Fruchart, M., Belarbi, K., Lamer, A., Dubois-Deruy, E., Lemdani, M., N’Guessan, A. L., Guinhouya, B. C., & Zitouni, D. (2022). Exploration of the core protein network under endometriosis symptomatology using a computational approach. In Frontiers in Endocrinology (Vol. 13). Frontiers Media SA. (DOI : https://doi.org/10.3389/fendo.2022.869053 )
Da Silva, C., Zitouni, D., Guinhouya, K. H., Storme, L., Hubert, H., Garat, A., Lamer, A., Lemdani, M., Tchédré, K. B., Gbenyo, K. F., Gasso, G., & Guinhouya, B. C. (2022). 414 – Approche computationnelle – Médicaments et voies biologiques dans les troubles du spectre autistique infantiles. In Revue d’Épidémiologie et de Santé Publique (Vol. 70, p. S158). Elsevier BV. (DOI: https://doi.org/10.1016/j.respe.2022.06.083 )
Patel, H., Patel, R., Zitouni, D., Guinhouya, B., Fruchart, M., & Lamer, A. (2022). Automated Twitter Extraction and Visual Analytics with
Dashboards: Development and First Experimentations. Studies in Health Technology and Informatics. IOS Press. (DOI: https://doi.org/10.3233/SHTI220562 )
M Messé, C Bernhard, M Mercier, Q Fuchs, S Foppolo, C Herold-Mende, I Namer, C Bund, M Elati, N Entz-Werlé, M Dontenwill, Heterogeneity and plasticity of integrin α5β1 expression in glioblastoma stem cells, Neuro-Oncology, Volume 23, Issue Supplement_2, September 2021, Page ii22. (DOI: https://doi.org/10.1093/neuonc/noab180.072 )
2021
A. Assi, W. Dhifli. Instance Matching in Knowledge Graphs through Random Walks and Semantics. Future Generation Computer Systems (Elsevier) 123:73-84, 2021. (DOI: https://doi.org10.1016/j.future.2021.04.015 )
Champion, M., Chiquet, J., Neuvial, P., Elati, M., Radvanyi, F., & Birmelé, E. (2021). Identification of deregulation mechanisms specific to cancer subtypes. Journal of Bioinformatics and Computational Biology, 19(01), 2140003. (DOI: https://doi.org/10.1142/S0219720021400035 )
Miagoux, Q., Singh, V., de Mézquita, D., Chaudru, V., Elati, M., Petit-Teixeira, E., & Niarakis, A. (2021). Inference of an integrative, executable network for rheumatoid arthritis combining data-driven machine learning approaches and a state-of-the-art mechanistic disease map. Journal of Personalized Medicine, 11(8), 785. (DOI : https://doi.org/10.3390/jpm11080785 )
C.E. Berbague, N.E.I. Karabadji, H. Seridi, P. Symeonidis, Y. Manolopoulos, W. Dhifli. An Overlapping Clustering Approach for Precision, Diversity and Novelty-Aware Recommendations. Expert Systems With Applications (Elsevier) 177:114917, 2021. (DOI: https://doi.org/10.1016/j.eswa.2021.114917 )
Geles K, Palumbo D, Sellitto A et al. WIND (Workflow for pIRNAs aNd beyonD): a strategy for in-depth analysis of small RNA-seq data.
F1000Research 2021, 10:1 (DOI : https://doi.org/10.12688/f1000research.27868.3 )
2020
W. Dhifli, N.E.I. Karabadji, M. Elati. Evolutionary Mining of Skyline Clusters of Attributed Graph Data. Information Sciences (Elsevier) 509:501-514, 2020. (DOI: https://doi.org/10.1016/j.ins.2018.09.053 )
Zerrouk, N., Miagoux, Q., Dispot, A., Elati, M., & Niarakis, A. (2020). Identification of putative master regulators in rheumatoid arthritis
synovial fibroblasts using gene expression data and network inference. Scientific reports, 10(1), 1-13. (DOI: https://doi.org/10.1038/s41598-020-73147-4 )
El Ati, Z., Machfar, H., Boussafa, H., Ati, N., Sioud, O. B. O., Zantour, B., … & Elati, M. (2020). Metabolic syndrome, malnutrition, and its associations with cardiovascular and all-cause mortality in hemodialysis patients: Follow-up for three years. Journal of Kidney Diseases and Transplantation, 31(1), 129. (DOI: https://doi.org/10.4103/1319-2442.279932 )
A. Assi, H. Mcheick, W. Dhifli. Data Linking Over RDF Knowledge Graphs: A Survey. Concurrency and Computation: Practice and Experience (Wiley) 32(19):e5746, 2020. (DOI: https://doi.org/10.1002/cpe.5746 )
2019
Coutant, A., Roper, K., Trejo-Banos, D., Bouthinon, D., Carpenter, M., Grzebyta, Santini, G., Soldano, H., Elati, M. # , Ramon, J. #, Rouveirol, C. #, Saldatova, L. #, Ross, K. #, Closed-loop cycles of experiment design, execution, and learning accelerate systems biology model development in yeast. Proceedings of the National Academy of Sciences (PNAS), 116(36), 18142-18147. 2019. # These authors co-supervised the study (EU Chist-Era AdaLab projet). (DOI: https://doi.org/10.1073/pnas.190054811 )
W. Dhifli, Julia Puig, Aurélien Dispot, M. Elati. Latent network-based representations for large-scale gene expression data analysis. BMC Bioinformatics 19(13):466, 2019. (DOI: https://doi.org10.1186/s12859-018-2481-y )
A. Assi, H. Mcheick, A. Karawash, W. Dhifli. Context-Aware Instance Matching Through Graph Embedding in Lexical Semantic Space.
Knowledge-Based Systems (Elsevier) 186:104925, 2019. (DOI: https://doi.org/10.1016/j.knosys.2019.104925 )
N.E.I. Karabadji, I. Khelf, H. Seridi, S. Aridhi, D. Remond, W. Dhifli. A Data Sampling and Attribute Selection Strategy for Improving Decision Tree Construction. Expert Systems With Applications (Elsevier) 129:84-96, 2019 (DOI: https://doi.org/10.1016/j.eswa.2019.03.052 )
T. K. Saha, A. Katebi, W. Dhifli, M. Al Hassan. Discovery of Functional Motifs from the Interface Region of Oligomeric Proteins using Frequent Subgraph Mining. IEEE/ACM Transactions on Computational Biology and Bioinformatics 16(5): 1537 – 1549, 2019 (DOI: https://doi.org/10.1109/TCBB.2017.2756879 )
R. Saidi, W. Dhifli, M. Maddouri, E. M. Nguifo. Efficiently mining recurrent substructures from protein 3D-structure graphs. Journal of Computational Biology (JCB)(Mary Ann Liebert, Inc.) 26(6):561–571, 2019. (DOI: https://doi.org/10.1089/cmb.2018.0171 )
A. Assi, H. Mcheick, W. Dhifli. BIGMat: A Distributed Affinity-Preserving Random Walk Strategy for Instance Matching on Knowledge Graphs. IEEE BigData: pages 1028-1033, 2019. (DOI: https://doi.org/10.1109/BigData47090.2019.9006348 )
2018
A. Lopez-Rincon, A. Tonda, M. Elati, O. Schwander, B. Piwowarski, P. Gallinari. Evolutionary Optimization of Convolutional Neural Networks for Cancer miRNA Biomarkers Classification. Applied Soft Computing, 65: 91-100, 2019. DOI: https://doi.org/10.1016/j.asoc.2017.12.036 )
Karabadji, N. E. I., Beldjoudi, S., Seridi, H., Aridhi, S., & Dhifli, W. (2018). Improving memory-based user collaborative filtering with evolutionary multi-objective optimization. Expert Systems with Applications, 98, 153-165 (DOI: https://doi.org/10.1016/j.eswa.2018.01.015 )
SOFTWARE
CoRegFlux : R Bioconductor package (Trejo el al., BMC Systems Biology2017, Coutant et al., PNAS 2019). Integrating transcriptional activity in genome-scale models of metabolism.
LatNet : R package (Dhifli el al., BMC Bioinformatics 2018). Latent network-based representations for large-scale gene expression data analysis.
CoRegNet : is an R bioconductor package, which enables learning of gene regulatory networks from transcriptome data and infers master regulators controlling the transition between phenotypes. (Download stats: https://bioconductor.org/packages/stats/bioc/CoRegNet/).
PEPPER : is a plugin cytoscape, which enables finding pathways connecting a protein set within a PPI-network using multi-objective optimization. Published in 2014 (Download stats: apps.cytoscape.org/download/stats/pepper/)
GREAT : The Genome REgulatory and Architecture Tools (GREAT). GREAT is a software suite of related and interconnected tools, currently able to perform systematic analyses of genome regularities (GREAT-Scan-Pattern) as well as improve TFBS prediction based on gene position information (GREATScan- PreCisIon).
Galaxy-X : An open-set multi-class classification method. Existing classification methods are designed to classify unknown instances within a set of previously known training classes. In open-set classification, the classifier may be confronted, during prediction, to observations that do not belong to any of the classes seen in training.
PROTNN : PROTNN is a function prediction approach for protein 3D-structures. PROTNN assigns to a query protein the function with the highest number of votes across the set of k nearest neighbor reference proteins based on a graph representation model and a pairwise similarity between vector embedding of both the query and the reference protein-graphs in structural and topological spaces.
PGR : An online repository of graphs representing all known protein 3D-structures. It aims at providing bioinformatics tools that facilitate the integration of graph theory techniques in the core of protein 3D-structure studies.
UnSubPatt : It is a feature selection approach for subgraphs. It selects a subset of representative subgraphs (called unsubstituted) from a given set of subgraphs based on label similarity defined in the domain knowledge.
TRS : Feature selection approach for subgraphs. It selects a small subset of topological representative subgraphs from a large set of subgraphs based on an embedding of subgraphs using a set of topological descriptors.
Ant-Motif : It is used to extract common substructures with an ant-like-shape (called ant-motifs) from a set of traceable graphs i.e., graphs having Hamiltonian path (eg, graphs of protein structures).
On-going theses
- PAWLAK Geoffrey (D3 in 2023), Explainable AI-based Lung cancer systems biology to decipher regulatory networks of oncogene addiction. Directors : Prof. Mohamed Elati / Dr David Tulasne
- LANERET Nathan (D2 in 2023), Systems biology of breast cancer metastasis. Directors : Prof. Robert-Alain Toillon/ Prof. Mohamed Elati
- HADJADJI Ines (D2 en 2023), Contributions to deep learning for genomics. Directors : DR. Nour Karabadji/Dr. Wajdi Dhifli
- YIN Liangwei (D1 in 2023), Bioinformatics and artificial intelligence approaches to assess the biological relevance of organoids and tumor spheroids. Directors : Dr. Christophe Battail/ Prof. Mohamed Elati
Defended theses
2020
- ASSI Ali, « Instance matching in knowledge graphs », Directeurs : H. Mcheick/W. Dhifli
2019
- TREBULLE Pauline, « Modélisation multi-échelles de réseaux biologiques pour l’ingénierie métabolique d’un châssis biotechnologique », Directeurs : Mohamed Elati/Jean-Marc Nicaud
2015
- NICOLLE Remy, « Regulatory networks driving bladder cancer », Directeurs : Mohamed Elati/François Radvanyi