Refining generative models for B-cells recombination and mutation.

Refining generative models for B-cells recombination and mutation. Inria is the French National Institute for Research in Digital Science, of which the Inria Côte d'Azur University Center is a part. With strong expertise in computer science and applied mathematics, the research projects of the Inria Côte d'Azur University Center cover all aspects of digital science and technology and generate innovation. Based mainly in Sophia Antipolis, but also in Nice and Montpellier, it brings together 47 research teams and nine support services. It is active in the fields of artificial intelligence, data science, IT system security, robotics, network engineering, natural risk prevention, ecological transition, digital biology, computational neuroscience, health data, and more. The Inria Center at Université Côte d'Azur is a major player in terms of scientific excellence, thanks to the results it has achieved and its collaborations at both European and international level. General Information Duration: 5-6 month Location: Marseille (on site) Internship start: January to April Internship stipend : ~670€ / month COMPO team The internship will take place within the COMPO team on the Timone Campus in Marseille. The ambition of the COMPO Inria‑Inserm joint project‑team is to develop novel mathematical models integrating data available in clinical oncology (from clinical trials and routine care), in order to provide decision‑making tools to oncologists. To achieve this, the team uniquely gathers mathematicians, pharmacologists and medical oncologists. It is integrated in the Center of Cancer Research of Marseille (Inserm U1068, CNRS UMR7258, Aix‑Marseille Université UM105, Institut Paoli‑Calmettes) and located in the La Timone Health Science campus of the University Hospitals of Marseille (AP‑HM), close to the INCa‑labeled center for early phase clinical trials (CLIP2). This joint project‑team, built on strong expertise in mathematical modelling, pharmacometrics and experimental and clinical oncology, is committed to develop novel methodologies combining mechanistic and statistical modelling to be ultimately applied at bedside. The work will take part along with the resumption of development of the IGoR software driven by several Inria senior software engineers (from Inria‑Sophia SED) through several coding sprints to which the trainee is strongly encouraged to attend. Context and Objectives of the Internship Motivation : The adaptive immune system comprises antigen‑specific lymphocytes, namely B and T lymphocytes. Each of these lymphocytes carries numerous copies of a unique receptor T cell receptors (TCRs) for T lymphocytes and B cell receptors (BCRs) for B lymphocytes. The nucleotide sequence of these receptors is not contained in the individual germline DNA but is randomly generated through the process of VDJ recombination. This stochastic germline DNA editing process combines a combinatorial choice among V, D and J gene families, together with random nucleotide deletions and insertions to generate an enormous diversity of T and B cells, each carrying their specific TCR and BCR. It is this randomly generated diverse set of lymphocytes, mostly created prior to birth, that constitutes the immune repertoire, and enables the specific recognition of virtually any pathogen throughout an individual’s lifetime. Owing to this complexity, building quantitative frameworks to extract meaningful information from high‑throughput Adaptive Immune Receptor Repertoire (AIRR) sequencing experiments is at the cornerstone of understanding the adaptive immune system in health and disease. Internship Projects : In the past we have successfully built mechanistic statistical generative models to the probability of generation (Pgen) of nucleotide sequences of V(D)J‑recombined nucleotide sequences and somatic mutations introduced by affinity maturation using the IGoR software. While the inferred models have been successfully applied to predict receptor sharing statistics, detect expanded clones or detect B‑cell lineages to name a few, they fall short of describing observed phenomena such as: Tandem D segments and longer insertion profiles: some experiments suggested that in a minority of recombination events two D‑gene segments may be included in the resulting sequence. Owing to the short length of D genes and random deletions occurring during recombination such tandem D insertions are difficult to tell apart from random nucleotide insertions. Using generative models of VDJ recombination allowing only for a single D gene segment we have previously quantitatively demonstrated that insertions alone were not sufficient to reproduce the number of sequences displaying tandem D compatible sequences. The development of generative models accommodating for such tandem D insertions would shed new light on the differences between T and B cell VDJ recombination while providing more accurate predictions for downstream use. Population‑level allelic variability in VDJ genes : researchers in computational immunology have long relied on curated lists of allelic variants e.g. from the IMGT database. However with the increased availability of repertoire sequencing data it has become evident that current sets are incomplete and unrepresentative of the degree of polymorphism and diversity in human and animal populations, while containing many spurious allele for a given individual. While some solutions have been proposed by the community to infer alleles from repertoire sequencing data, these solutions only address V gene polymorphism. Adapting such approaches to IGoR’s probabilistic framework, coupled with a previously developped haplotype inference strategy, could allow inference of germline variants even in the short D and J genes. Keywords: Adaptive immune system; B cells; VDJ recombination; Somatic Hypermutations (SHMs); Mutational Signatures; Generative Models; statistical data analysis Main activities (5 maximum) : Literature review & hypothesis framing Data analysis & Visualization Communication of the results in team meetings Write reports Desired Skills We are looking for a motivated candidate with skills in probability, statistics, and scientific computing coming from a bio‑informatics, applied mathematics or physics background. Knowledge of biology is not required but valued, and an interest in applications to immunology and cancer is expected. Programming skills in Python and C/C++ languages are necessary. Avantages Partial reimbursement of public transport costs Leave: 7 weeks of annual leave + 10 extra days off due to RTT (statutory reduction in working hours) + possibility of exceptional leave (sick children, moving home, etc.) Possibility of teleworking (after 6 months of employment) and flexible organization of working hours Professional equipment available (videoconferencing, loan of computer equipment, etc.) Social, cultural and sports events and activities Access to vocational training Social security coverage #J-18808-Ljbffr