Systems Biology lab group meeting will take place on
Monday, 10/25/2021
10:00am
Virtual
Presenter: Andreas Tjärnberg
Title: Gene regulation network inference through latent factor activity and interpretable deep models
-------------------------
Gene regulation is dependent on context and is non-linear. Many regulatory components can not be directly measured in large genome-wide screens and are either ignored or inferred when modeling gene regulation. Therefore most inference methods are forced to model covariance of regulatory genes on their targets as a proxy for causal regulatory influence. This in turn complicates validation and reuse of predictive models. To disentangle covariance and casual influence require orthogonal evidence and is often done with the use of latent features such as the activity of transcription factors (TFA).
Monday, 10/25/2021
10:00am
Virtual
Presenter: Andreas Tjärnberg
Title: Gene regulation network inference through latent factor activity and interpretable deep models
-------------------------
Gene regulation is dependent on context and is non-linear. Many regulatory components can not be directly measured in large genome-wide screens and are either ignored or inferred when modeling gene regulation. Therefore most inference methods are forced to model covariance of regulatory genes on their targets as a proxy for causal regulatory influence. This in turn complicates validation and reuse of predictive models. To disentangle covariance and casual influence require orthogonal evidence and is often done with the use of latent features such as the activity of transcription factors (TFA).
Gene regulatory network inference using latent features such as transcription factor activity can be built into a single framework while maintaining interpretable parameter estimates using the paradigm of deep learning.
In this presentation, I will present my attempts at achieving such a model. Demonstrating these results in Yeast single-cell sequencing data.
