Bravo à Camille Peneau et Pablo Rodriguez qui ont remporté les prix "poster" décernés par la SFBI !
Le parcours de Camille :
Au cours de mes études d'ingénieur à l'Ecole Centrale Paris, je me suis progressivement orientée vers la recherche en biologie grâce à plusieurs stages en France, au Canada et aux Etats-Unis et à un M2 en bio-physique. J'ai ensuite débuté ma thèse en 2016 au Centre de recherche des Cordeliers à Paris au sein de l'équipe Génomique des tumeurs solides dirigée par le Pr Jessica Zucman-Rossi. Mon projet de recherche est à l'interface entre la biologie expérimentale et la bio-informatique et repose sur l'étude des génomes du virus de l'Hépatite B présents dans les tumeurs du foie.
Le résumé de son poster :
Hepatitis B virus (HBV) is the leading risk factor for hepatocellular carcinoma (HCC) occurrence, the third cause of cancer death worldwide. HBV is a DNA virus that could integrate in human DNA and promote cell transformation by insertional mutagenesis. As HBV integrations occur early during infection, it is a key factor reflected in the genetic landscape of HCC. However, up to now, identifications of HBV insertions in HCC have been mainly performed in Asian populations. Our project aimed to characterize HBV-related insertional mutagenesis in a large cohort of HCC from patients with European or African origin. We performed viral capture and next-generation sequencing on 220 HCC and their normal liver counterparts from 180 HBV-positive patients. We set up an in-house pipeline of analysis to characterize precisely the integration sites and extract the integrated sequences. We found that HBV integration can occur multiple times in the same cellular clone and the locus of TERT is the main recurrent hotspot of insertion in tumors. The in-silico reconstruction of integrated sequences revealed the existence of structural rearrangements in the viral sequence as in the human genome around the integration breakpoints. This study provides a global view of the landscape of HBV integrations in European and African populations, by characterizing the different viral forms and sequences in tumors and non-tumor liver tissues.
Le parcours de Pablo :
After finishing a PhD in Computer Science & Artificial Intelligence in the University of Santiago de Compostela (Spain), I reoriented my research career to bring together two of my passions: biology and informatics. I was fortunate enough to find a very interesting research project at INRA Toxalim (MetExplore) in Toulouse, which I'm currently working on as a postdoctoral researcher. In this project, I am in charge of the development of the computational methods to discover novel metabolic dysregulations in cancer due to mutations in the p53 gene, one of the most frequently mutated genes in tumours. We aim to use these computational models to predict new potential biomarkers of cancer development that are relevant for Li-Fraumeni patients that carry germline mutations in this gene.
Le résumé de son poster :
Tissue specific constraint-based modelling approaches have proven useful as automatic ways of extracting and analysing metabolic networks that capture the different metabolic states of cells. These methods integrate different sources of information such as stoichiometry, transcriptomics or metabolomics data that constrain the space of possible metabolic networks. This process is usually done by searching for a metabolic network that minimizes an objective function measuring the discrepancy between the observed data and the model. However, current methods usually extract one single optimal network from which all the subsequent analysis is derived. However, this solution may not be unique, meaning that the observed data can be explained by a set of equally good metabolic networks. Ignoring this variability may lead to incorrect or incomplete explanations and bias the interpretation. In order to analyse this effect, we developed an extension of iMAT, a method for extracting tissue-specific networks from transcriptomics data, to enumerate different alternative optimal networks generated from the same data. Our study highlights the importance of analysing the space of alternative optimal solutions as a way to reduce potential bias in the interpretation of data using constraint-based modelling approaches.