Math Modeling Meets Machine Learning

Focused on facilitating the iterative refinement of ground truth in the bio-sciences through smarter collaboration and co-opetitions, our innovation challenges are within the space of Systems Biology, and more importantly, Structural Systems Biology.

The core objective of these innovation challenges are to merge and integrate current tools and techniques in Logic, traditional Mathematical Modeling, and newer Machine Learning approaches with an aim of developing better Systems and Structural Systems Biology Models. Sub-tasks will include but are not limited to, data augmentation, learning of biologically consistent constraints, better interpretability and interoperability of models, and generally, more robust end to end hybrid models. 

Although open to change, or current innovation challenges are designed based on recent advances in Machine Learning, coupled with the work on Constraint-Based Modeling at Dr. Bernhard Palsson’s Lab at UCSD, and work on Data-Driven Sparse Identification of Nonlinear Dynamics Of Biological Networks at Steven Brunton’s lab at University of Washington.

What has now become an intensely competitive territory will inevitably need to be redirected into a more collaborative space for the continued engagement of all the brilliant minds coming into and already in the field. This includes tackling certain problems such as assembly order of multimeric complexes. 

Embed, encode, attend & predict.