My recent interest has been shifting to scientific applications of AI/ML, especially to biology where there is enormous amount of data with hidden cross-correlations and causations waiting to be discovered. I have been undertaking serious literature survey on foundational ML and its applications, and looking to make contributions. I'll update this page as things happen.
Embedded Files
Genetic Tool to Cell Type Bayesian Mapper (GT2CT)
Allen Institute at Seattle announced a data challenge called MapMySections. The challenge, as stated, was to map serial two-photon tomography images from the Genetic Tools Atlas to spatial transcriptomic cell types in the ABC Atlas of the mouse brain.
With Chaitali Khan at the NIH, I developed GT2CT -- an ML-based tool kit with firm Bayesian foundations to perform this mapping. As amateurs in this field, we were quite surprised when we were chosen as one of the winning teams of this challenge.
If you are a neurobiologist, I will be delighted to hear your thoughts on the utility of this tool and avenues for improvement.
Page updated
Google Sites
Report abuse