Helene Ruffieux

I am Postdoctoral Research Associate working with Prof. Sylvia Richardson and Dr Leonardo Bottolo on Bayesian modelling for sparse regression problems. My work involves designing flexible hierarchical approaches to infer complex biological structures from genome-wide association and molecular quantitative trait locus data, with a special emphasis on the modelling of pleiotropic signals. I am also interested in enhancing machine learning algorithms – in particular variational inference – in a context-specific fashion, in order to better manage the inherent tension between expressive joint modelling and scalability of inference.

I hold a Ph.D. in Mathematics from Ecole Polytechnique Fédérale de Lausanne (EPFL), jointly advised by Prof. Anthony C. Davison and Dr Jörg Hager. Before and during my doctoral studies, I have also worked for the Swiss Institute of Bioinformatics and Nestlé Research alongside geneticists and computational biologists, which reinforced my initial interest in developing methodologies to improve basic biological research and its translation into medical care.

• H. Ruffieux, A. C. Davison, J. Hager, and I. Irincheeva. (2017) Efficient inference for genetic association studies with multiple outcomes Biostatistics 18: 618–636
• H. Ruffieux, A. C. Davison, J. Hager, J. Inshaw, B. Fairfax, S. Richardson, and L. Bottolo (2018) A global-local approach for detecting hotspots in multiple response regression arXiv:1811.03334
• A. Valsesia, Q.-P. Wang, N. Gheldof, J. Carayol, H. Ruffieux et al. (2019) Genome-wide gene-based analyses of weight loss interventions identify a potential role for NKX6.3 in metabolism Nature Communications 10: 540
• H. Ruffieux, J. Carayol, M. E. Harper, R. Dent, W. H. M. Saris, A. Astrup, A. C. Davison, J. Hager, and A. Valsesia (2019) A Bayesian joint pQTL study sheds light on the genetic architecture of obesity biorxiv:10.1101/524405

• ATLASQTL (fast global-local hotspot QTL detection)
• ECHOSEQ (replication and simulation of molecular and clinical data)
• LOCUS (large-scale variational inference for variable selection in sparse multiple-response regression)