Integration of network information and gene expression data.
Exploratory approach and expected results. Our goal is to put RNAseq measurements in the context of the molecular networks that govern cells activity. Large transcriptome profiles for different tissues enable researchers to seek tissue-specific patterns. This can be done via the identification of individual genes. However, biological processes are driven by the many interactions of their constituting entities. The interaction networks gather all the known interactions, for all conditions. From this “generic multiplex” network, we aim to identify active sub-clusters under specific conditions by contextualising the network.
This idea comes within the scope of untangling the complexity of some genotype-phenotype relationships by identifying underlying causal (communities within) regulatory networks. To tackle noise, high-dimensionality and heterogeneity, we will use probabilistic graphical models (PGM) to contextualise relationships between genetic mutation, gene expressions and phenotypes of interest. For example, Markov random fields, a class of PGM, can integrate multiplex networks -a prior in statistical terms- with tissue or cell-specific data. We will map the landscape of cellular and tissue perturbations and identify subparts of the integral genetic circuitry specific to these conditions.
Ultimately, we would like to transpose the developped methodology to the study of rare monogenic diseases (MD), characterised by mutations in single genes triggering devastating health disorders for affected individuals. MDs display largely unexplained variability in symptoms, causing many patients to remain undiagnosed, with almost no existing treatment (sample sizes can be very low) could be treated as specific conditions with a commmon triggering factor. Network contextualisation in this setting is a first step into personalised medicine.
Role of the candidate and profile:
The successful candidate will be responsible for the study of the data integration technique to jointly analyse multiplex networks and transcriptomics data and the associated implementation, e.g. in the form of an R package. A report will also be expected to be produced at the end of the internship.
Some mathematical knowledge and computational (Python and/or C and/or R) skills with a real taste for multi-disciplinary collaborations (bioinformatics, medical health) are needed for this project. Masters or 3rd year/gap year/end of engineering school curriculum with 1st class or 2nd class upper honours (a solid ‘Assez Bien’ merit for the L3/M1 in the French system).
A stipend to cover travel costs and living expenses in Palmerston North (NZ) is available to the right candidate for the duration of the internship (6 month). The supervision team will include mathematicians/statisticians/bioinformaticians from Massey University (NZ) and Aix-Marseille* University (France).
Starting date is negotiable but the sooner the better.
To apply, please send your CV (including at least two refrerences) and a motivation letter (1 page max).