M2 Genomics Informatics and Mathematics for Health and Environment

Fariza Tahi, PR UEVE.

A project will be proposed and evaluated, as well as a study of scientific publications.

The goal of the course is firstly to define the non-coding RNAs (ncRNAs), including the microRNAs (miRNAs), and then to present their involvement in diverse human diseases, as well as their use in genetic therapy. The second part of the course will present the bioinformatics methods existing for their prediction, and the challenges that remain, particularly for identification at large scale and for big data.

The course will cover the ifferent bioinformatics approaches existing for RNA prediction
- RNA secondary structure prediction
- RNA complexes structure prediction
- RNA 3D structure prediction
- microRNA prediction. Examples of methods and software : advantages and disadvantages.
- piRNA prediction.
- Prediction of all classes of ncRNAs.
- Methods for non-coding RNA classification.

Basics in molecular biology and in functional RNA Basics in sequence algorithmic and machine learning.

Platon L, Zehraoui F, Bendahmane A, Tahi F. IRSOM, a reliable identifier of ncRNAs based on supervised self-organizing maps with rejection. Bioinformatics. 2018 Sep 1;34(17):i620-i628.

Legendre A, Angel E, Tahi F. Bi-objective integer programming for RNA secondary structure prediction with pseudoknots. BMC Bioinformatics. 2018 Jan 15;19(1):13.

Boucheham A, Sommard V, Zehraoui F, Boualem A, Batouche M, Bendahmane A, Israeli D, Tahi F. NCBINCBI Logo IpiRId: Integrative approach for piRNA prediction using genomic and epigenomic data. PLoS One. 2017 Jun 16;12(6):e0179787.

Legendre A, Angel E

Septembre - Octobre - Novembre.

EVRY

Franck Delaplace, PR UEVE Sergiu Ivanov, Associate Professor, UEVE Anna Niaraki, Associate Professor, UEVE.

Contents of the UE: 1.Introduction to Systems Biology and Network Medicine 2.Foundations a.Topological analysis i.Graphs: definition and fundamental algorithms ii.Graph measures iii.Case studies iv.Random networks study b.Discrete dynamics i.Boolean networks ii.Petri nets 3.Applications to systems biology & tools a.Introduction i.Data formats : SBML, SBGN, etc. b.Types of biological networks i.Types and functional roles ii.Databases iii.Types and data acquisition c.Integration of data into networks i.Network inference ii.Functional integration (Cytoscape + REACTOME) d.Modelling of network dynamics i.Modelling principles ii.Case studies and software tools (GINsim).

The objective of this course is to provide the fundamental background combining theoretical, computational and biological knowledge for this challenge. The course introduces the theoretical frameworks and the computational methods applied to Systems Biology, with a specific focus on networks. The content of the course is divided in three major parts:
The first part will be dedicated to the fundamental knowledge of networks, and in-depth topological analysis of complex networks. Different types of networks, case studies and applications in biology will be discussed.
In the second part, emphasis will be on dynamical networks and on how discrete based mathematical formalisms can be used to describe their dynamics and delineate their properties. Two modelling approaches will be presented, Boolean networks and Petri nets, with concrete examples in biology.
In the third part, the students will be introduced to the standards of network and model description commonly used in Systems Biology. Then, different types of intracellular networks will be analyzed in detail, along with a thorough presentation of available resources such as model repositories and databases. Bottom up and top down approaches for network construction and inference will be discussed and different data types will be presented. Lastly, the students will learn about functional data integration and limitations.

The prerequisites for the course are a solid background in molecular and cellular biology, and a confirmed grasp of computer programming. The students are also assumed to have basic knowledge of discrete mathematics.

Septembre - Octobre - Novembre.

EVRY

Marie-Anne Debily, Associate professo UEVE Bérengère Genin, Director of Bioinformatics, Integragen Researchers of Genopole® campus and Gustave Roussy Cancer center.

The course consists in lectures, analysis of genomic studies based on several publications during tutorials and illustrations by research seminars given by Genopole® campus and Gustave Roussy Cancer center experts that will highlight the contribution of these global approaches to the understanding of cellular processes and human complex diseases. Oral presentations of scientific research will be held by the students in a seminar format during the course.

This course provides a thorough knowledge of genome-wide studies from experimental design to integrative data analysis: Explore genomic variations in both health and disease to understand gene function and regulation of expression, decipher pathogenesis mechanisms, identify therapeutic opportunities and treatments in the context of precision medicine.

-Enumerate and differentiate large-scale sequencing technology driven-approaches at different levels, i.e. genomics, transcriptomics, epigenetic landscapes.
-Determine which technology to use among a broad spectrum of functional genomics methods to address specific biological questions
-Precise the main classification methods œfor patient stratification and identification of co-regulated genes
-Interpret results of large-scale experimental datasets in a scientifically stringent manner
-Critically examine research publications dealing with functional genomics
-Understand how to identify altered signaling pathways, biological process and biomarkers from genes list by functional annotation tools,
-Specify the advantages of single-cell approaches regarding the dynamic of transcriptome, tissue complexity or polyclonal sample content.

Septembre - Octobre - Novembre.

EVRY

Carène RIZZON, Associate Professor, UEVE Nathalie BOUDET, Associate Professor, UEVE.

Evolutionary questions will be addressed from raw data (sequenced genomes, proteomes, ...) through practical teaching and projects.

The course will address the structure and the evolution of prokaryotes and eukaryotes genomes and how comparative genomics (through bioinformatic) is a powerful tool to better understand them.
Schedule:
- Evolution of genomes (structure, duplicated genes, transposable elements, ...)
- Bioinformatics tools and methods to compare genomes (similarity search, gene network analysis, phylogeny, functional annotation, Ka and Ks parameters …).

Knowledge on evolution mechanisms and structure of genomes Knowledge on script language (Shell, awk, Perl, Python...) Knowledge on usual alignement tools (BLAST, CLUSTALW...).

Septembre - Octobre - Novembre.

EVRY

•Abdelghani SGHIR : PR, Université d’Evry-Paris-Saclay, Institut de Biologie François Jacob, CEA-Genoscope •Eric Pelletier : Institut de Biologie François Jacob, CEA-Genoscope •Amine Madoui : Institut de Biologie François Jacob, CEA-Genoscope •Catherine Dauga : Institut Pasteur •Joël Doré, INRA : INRA, Jouy-En-Josas.

This course is organized on a week, it takes place during the first semester. It includes four days of lectures by nationally and internationally renowned scientists in the field of metagenomics and multiomics. The lectures will be followed by a day of tutorials in computer room. Course slides in pdf format of each conference will be available on E-Campus. The last half a day will be dedicated to students oral presentation of articles and projects, followed by questions from the audience.

The objectives of this module are to provide the students an insight of complex microbial communities in terms of structure, phylogeny and metabolic diversity using computational biology and statistical tools for analysis of high throughput sequencing of metagenomes.
Metagenomics or ecogenomics is based on genomic analysis of complex microbial communities extracted from different environments such as fresh water or wastewater, marine water, digestive tract, soil, etc. Ecogenomics potentially enable a better understanding of the composition of the different microbial communities, their interactions with each other and their environments without relying on the cultivation techniques. Bioinformaticians will review the different approaches of sequence analysis of a metagenome: alignment, assembly, identification, phylogeny and functional gene screening and annotations. Exploring metagenomes for gene discovery to complete our knowledge of microbial metabolism or to search for novel enzymes with potential use in different area such as biocatalysts for green chemistry. Hence Industrial and biotechnological applications of metagenomics will be addressed.

Basic skills in : General Microbiology and Environmental microbiology, physicochemical characterization. Biochemistry and microbial metabolism. Cellular and molecular biology. Functional organization of the cell. Microbial genomics, Molecular phylogeny Biostatistics and computational Biology.

Metagenomics, Theory and Applications. Diana Marco (WWW.caister.com) Pelletier, E., A. Kreimeyer, et al. (2008). Candidatus Cloacamonas acidaminovorans: genome sequence reconstruction provides a first glimpse of a new bacterial division. J Bacteriol 190(7): 2572-2579. Wooley, J. C., A. Godzik, et al. (2010). A primer on metagenomics." PLoS Comput Biol 6(2): 1000667. Julia Hassa et al., 2018. Metagenome, metatranscriptome, and metaproteome approaches unraveled compositions and functional relationships of microbial communities residing in biogas plants. Applied Microbiology and Biotechnology

Novembre - Décembre.

EVRY

Farida Zehraoui, Associate Professor, UEVE Other researchers from UEVE working on IA and machine learning.

The aim of the course is to provide a good foundation on deep learning as well as practical, hands-on training to enable the student to implement his own deep learning algorithms.

Personalized medicine has the potential to profoundly improve the practice of medicine. Genomics is already used to guide clinical decision making and its contribution is likely to increase. ”Omics” technologies : genomics (DNA sequencing), transcriptomics (microarrays), proteomics, have significantly altered the scale of data and can generate massive amounts of genomic data on patients. These data can cover all the mechanisms involved in the variations that occur in cellular networks that influence the functioning of organic systems in humans.
The goal of the course is to provide students basis on deep learning algorithms that can use Omics data in addition to physiological features of patients to make diagnosis, prognosis, prediction of personalized treatment, etc. The obtained deep learning models can be very helpful for physicians and / or patients.

The course will cover the following topics:
Perceptron and Deep Feedforward Networks
Regularization for Deep Learning
Optimization for Training Deep Models?Autoencoders
Convolutional Networks
Recurrent and Recursive Nets
Interpretation / explanation of deep learning models.

Familiarity with basic statistics and basic machine learning algorithms is expected.

- Goodfellow, Bengio, Courville. Deep Learning, MIT press, 2016. http://www.deeplearningbook.org/ - Montavon, Grégoire, Wojciech Samek, and Klaus-Robert Müller. “Methods for Interpreting and Understanding Deep Neural Networks.” Digital Signal Processing.

Septembre - Octobre - Novembre.

EVRY

Christophe Ambroise, PR UEVE Jean-Christophe Janodet, PR UEVE Marie-Hélène Mucchielli-Giorgi, PR UEVE.

This UE is composed of three parts which revolve around the quantitative analysis of gene expression : 1. A first part on the statistical methods for gene expression data analysis 2. A second part on the inference of regulatory networks. 3. a last part on Bayesian networks Each part will be composed of a part dedictaed to course and a part dedictaed to Practical Works. All Practical works will be done from the same dataset.

- Microarray and RNA-seq technologies
- gene expression Data : large Data with high fluctuations
- statistical Methods for differential analysis of gene expression Data
- clustering and statistical Methods for coexpressed genes Identification
- statistical Methods for inference of gene regulatory network
- bayesian reasonning : Principles and difficulties
- definition and properties of The bayesian Networks
- exact and approximate inference
- designing and learning A network.

Basic Knowledge in Probability, Descriptive Statistics, Sampling and Hypothesis Testing, R programming.

Septembre - Octobre - Novembre.

EVRY

Fariza Tahi, PR UEVE.

This module aims to consolidate the professional project of our students, giving them a broad view and entrepreneurial culture, by offering conferences series, meetings and debates with professionals and researchers in the areas of bioinformatics and biotechnology.
Seminar series will be given by a broad range of bioinformatics, biotechnology and sociology experts.

Septembre - Octobre - Novembre.

EVRY

Fariza Tahi , PR UEVE Educational team of M2 GENIOMHE to tutor the internship.

The aim of this UE is to propose to the students bioinformatics projects on the whole of first semester on a research topic proposed and supervised by one or several researchers or professors of the master. The projects will begin by a state-of-art, allowing the students improving their knowledge in the field they have chosen. They will be then evaluated by a report and an oral presentation. The presentations will be organized as a seminar or a colloquium day.

Septembre - Octobre - Novembre - Décembre - Janvier.

EVRY

Patrice Lucas, PAST UEVE, CEA Bruyères-Les-Chatelles, Responsable opérationnel du Très grand centre de calcul du CEA Laurent Breda, PAST UEVE Cyrille Pauthenier, Ph.D., Président Abolis Biotechnologies, SAS,.

In this course we will explore the lifecycle (design/build/run) of a health application and constraints it has to overcome to be successful and to be accepted.
- The first part of the course will focus how to design and build a scalable application to launch or use many external components at the same time to be efficient. This design must be flexible to accept, to grow up, with all new breakthrough in Health.
- We will see the Domain Driven Design and the software Hexagonal Architecture to design a scalable application. And all the automatization to build a robust and reactive application.
- Feedback and theory on the application of the agile framework for the construction, management and maintainability of a code developed in team and in the long term.

The second part will provide minimum theoretical foundations for developing algorithms in a context of massively distributed architectures. This will include the different types of parallel architectures (including GPU), the parallel programming models. Then will be shown aspects of concurrent programming and distributed programmation, and programmation with GPUs.
The theory is then put into practice by implementing different parallel and distributed algorithms in C / C ++ using common tools such as POSIX threads, OpenMP and MPI.

Basics in programming.

Septembre - Octobre - Novembre - Décembre.

EVRY

Eric Angel, PR UEVE Franck Delaplace, PR UEVE.

Written exam.

Algorithmic techniques useful when dealing with NP-hard combinatorial optimization problems.

- approximation algorithms (vertex cover problem)
- integer programming
- randomized algorithms
- multicriteria optimization (Pareto front)
- metaheuristics (genetic algorithms, tabu search).

Basic knowledge in graphs, computational complexity and algorithmics.

Approximation algorithms, Vazirani V., Springer 2003.

Septembre - Octobre - Novembre.

EVRY

Nazim Agoulmine, PR UEVE.

Tutorials and Lectures with practical examples.

Understanding biological processes requires storing, accessing and processing many complex and more and more large scale datasets. These data are usually heterogeneous generated by humans, machines (imaging systems, sequencing systems, etc.) or by programs (simulations tools). In parallel, computer science has made significant progress in developing new methods, models and technologies to storing, integrating and processing big heterogeneous data in the Cloud Computing. It is of the benefit of modern biology to be able to use these concepts and tools to achieve its need in term of integration and processing of large amounts of heterogeneous data, with different formats and semantics, and manipulate them through complex workflows.

Programming langage Algorithmic Databases.

1) Cloud Computing for Science and Engineering, Ian Foster and Dennis B. Gannon, The MIT Press, Cambridge, Massachusetts, London, England © 2017 Massachusetts Institute of Technology 2) NoSQL Data Models: Trends and Challenge, Olivier Pivert John Wiley & Sons, 14 août 2018 3) BIG DATA AND HADOOP, Mayank Bhusan, BPB Publications, 2 juin 2018 - 333 pages.

Septembre - Octobre - Novembre.

EVRY

Fariza Tahi, PR UEVE Educational team of M2 GENIOMHE to tutor the internship.

Internship of 5 to 6 months in a research laboratory or entreprise.

Janvier - Février - Mars - Avril - Mai - Juin - Juillet.

EVRY