M2 - internship in deep learning for genomic at Ecole Normale Supérieure

Type de poste
Dates
Durée du poste
Contrat renouvelable
Contrat non renouvelable
Date de prise de fonction
Date de fin de validité de l'annonce
Localisation
Adresse

<p>Ecole Normale Supérieure 46, rue d&rsquo;Ulm in Paris, France</p>
Paris
France

Contacts
Auguste Genovesio
Email du/des contacts
auguste.genovesio@ens.fr
Description

<p>The M2 internship will take place at the Computational Bioimaging and Bioinformatics team at IBENS, Ecole Normale Supérieure 46, rue d&rsquo;Ulm in Paris, France. The research team is composed of computer and data scientists that develop algorithms and novel approaches to tackle concrete biological questions through large scale image and data analyses and modeling.</p><p>Biology and medicine have recently been identified as the fields where the amount of growing data may soon become the largest, above astronomy, above particle physics, above chemistry and above social media. This places research in life science as a place of choice to develop novel AI methods.</p><p>Genes produces proteins that are the fundamental building blocks of biological systems generated over millions of years of evolution by natural selection of random mutations in DNA. However, we are reduced at studying the subset of proteins that still exist today. Protein engineering shows increasing success at producing interesting variations of natural proteins, but remains an incremental approach preventing the emergence of radically novel biological function. The M2 internship project will be part of a wider project with an overall goal to develop a deep learning model capable of generating artificial gene coding sequences. To date, the team has succeeded in developing a deep neural network models that efficiently learnt latent rules underlying protein coding sequences (CDS). This system should in principle allow for identification of CDS in newly sequenced genomes.</p><p>The work of the M2 internship will consist in improving the CDS prediction model that was so far obtained. Depending on the student background, the first part of the project will consist in getting familiar with the deep learning theory, application and implementation framework as well as with the biological objects involved: DNA, protein coding sequence, exons, pseudo-genes, etc. The second part of the project will consist in raising the accuracy of the detection model obtained so far by efficient neural architecture search (ENAS) and demonstrate its broad applicability. The last part of the internship will be dedicated to the assembly of a complementary model that will be dedicated to the precise identification of intro-exon junctions in order to finely delineate CDS.</p><p>This project will be performed in collaboration with the Genome and Organization Dynamics (DYOGEN) team at IBENS.</p><p>Candidate profile: M2 in computer science, data science, machine/deep learning or bioinformatics. A basic knowledge in biology is welcome but not mandatory.</p><p>Applicants should send a CV, a motivation letter and contact details to Auguste Genovesio, <a href="mailto:auguste.genovesio@ens.fr">auguste.genovesio@ens.fr</a></p><p>Team web page:&nbsp;https://www.ibens.ens.fr/spip.php?rubrique47&nbsp;</p><p>&nbsp;</p><br/…;
Laboratoire: Computational Bioimaging and Bioinformatics team at IBENS, Ecole Normale Supérieure