Large-scale atlases of organs in a healthy state are soon going to be available, in particular, within the Human Cell Atlas. This is a significant step in better understanding cells, tissues and organs in healthy state and provides a reference when diagnosing, monitoring, and treating disease. However, due to the sheer number of possible combinations of treatment and disease conditions, expanding these data to characterize disease and disease treatment in traditional life science laboratories is labor intensive and costly and, hence, not scalable.
Accurately modeling cellular response to perturbations (e.g. disease, compounds, genetic interventions) is a central goal of computational biology. Although models based on statistical and mechanistic approaches exist, no machine-learning based solution viable for unobserved, high-dimensional phenomena has yet been available. In addition, scGen is the first tool that predicts cellular response out-of-sample. This means that scGen, if trained on data that capture the effect of perturbations for a given system, is able to make reliable predictions for a different system. "For the first time, we have the opportunity to use data generated in one model system such as mouse and use the data to predict disease or therapy response in human patients," said Mohammad Lotfollahi, PhD student (Helmholtz Zentrum München and Technische Universität München).
scGen is a generative deep learning model that leverages ideas from image, sequence and language processing, and, for the first time, applies these ideas to model the behavior of a cell in silico. The next step for the team concerns the improving scGen to a fully data-driven formulation, increasing its predictive power to enable the study of combinations of perturbations. "We can now start optimizing scGen to answer more and more complex questions about diseases," said Alex Wolf, Team Leader, and Fabian Theis, Director of the Institute of Computational Biology and Chair of Mathematical Modeling of Biological Systems at Technische Universität München.
The Institute of Computational Biology (ICB) develops and applies methods for the model-based description of biological systems, using a data-driven approach by integrating information on multiple scales ranging from single-cell time series to large-scale omics. Given the fast technological advances in molecular biology, the aim is to provide and collaboratively apply innovative tools with experimental groups in order to jointly advance the understanding and treatment of common human diseases. http://www.helmholtz-muenchen.de/icb
As German Research Center for Environmental Health, Helmholtz Zentrum München pursues the goal of developing personalized medical approaches for the prevention and therapy of major common diseases such as diabetes mellitus and lung diseases. To achieve this, it investigates the interaction of genetics, environmental factors and lifestyle. The Helmholtz Zentrum München has about 2,300 staff members and is headquartered in Neuherberg in the north of Munich. Helmholtz Zentrum München is a member of the Helmholtz Association, a community of 18 scientific-technical and medical-biological research centers with a total of about 37,000 staff members. https://www.helmholtz-muenchen.de/en/index.html
Original publication: Lotfollahi, M. et al. (2019): scGen predicts single-cell perturbation responses. Nature Methods, DOI: 10.1038/s41592-019-0494-8 https://www.nature.com/articles/s41592-019-0494-8
Related Articles: Clear sight in the data fog with PAGA
Contact for the media: Department of Communication, Helmholtz Zentrum München - German Research Center for Environmental Health, Ingolstädter Landstr. 1, 85764 Neuherberg - Tel. +49 89 3187 2986 - E-mail: email@example.com