Topic
Gene isoforms play key roles in disease etiology and outcome. Despite that gene isoforms are involved in most biological processes, they are commonly ignored because we lack tools to understand them. In this project we will build new large-scale machine learning tools to extract gene isoform information from commonly existing single-cell RNA-seq data.
The approach we consider is based on variational autoencoders and sequence models. The challenges lie in the data being extremely large, sparse, and noisy. The project will open doors into the currently most rapidly expanding field in biology.
Working tasks
You will carry out formulation, implementation and testing of machine learning approaches to model and understand single-cell RNAseq data. You will present your findings, write manuscripts, help supervising students, and actively contribute to a collegial lab culture.
You will be based in the lab of Johan Henriksson (www.henlab.org) in collaboration with the machine learning group of Associate Professor Tommy Löfstedt. The Henriksson lab is based at the Department of Molecular Biology and is part of The Laboratory of Molecular Infection Medicine Sweden (MIMS, www.mims.umu.se), which is the Swedish node within the Nordic EMBL Partnership for Molecular Medicine. The project is run in close collaboration with Tommy Löfstedt, docent and associate professor and head of the machine learning group at the Department of Computing Science.
Eligibility
A person who has been awarded a doctorate in relevant subject area or a foreign qualification deemed to be the equivalent of a doctorate qualifies for employment as a postdoctoral fellow. This eligibility requirement must be met no later than the time at which the appointment decision is made.
Postdoctoral fellows who are to teach or supervise must have taken relevant courses in teaching and learning in higher education.
You should be highly motivated and able to work productively in a team as well as independently. Excellent communication skills for interacting effectively with senior colleagues and peers are required. Proficiency in written and spoken English also required. Great emphasis will be placed on personal suitability.
A successful candidate should be familiar with Python and at least one framework for deep learning (e.g. PyTorch or TensorFlow/Keras). The candidate should have documented experience in machine learning or mathematical/statistical modelling.
Application