Programming Living Cells through Targeted Machine Learning - Databricks

Programming Living Cells through Targeted Machine Learning

Over the past decade, there has been a rise in the usage of biologics: biologically derived therapeutics to treat a breadth of human diseases including cancer and rare genetic diseases. The growth of treatment modalities such as antibodies, viral vectors and cell therapeutics has outpaced classic methods such as small organic molecule development. Despite this growth, methodologies for intelligently designing biological therapeutics are arcane; it can take years of research to discover a single genetic design that produces a functional therapeutic. In this talk we will walk through advances in experimental data collection and machine learning algorithms to accelerate the design of living cells to produce therapeutics. Specific problems such as promoter sequence design, codon optimization and genetic circuit design will be discussed in the context of machine learning solutions.

« back
About Joe Isaacson


Joe is the vice president of engineering at Asimov, a Cambridge, MA based company with the mission to radically advance humanity's ability to design living systems. Asimov researches machine learning methods to automate the design of living systems and applies these methods to facilitate the production of next generation therapeutics. Previously, Joe lead machine learning efforts building ad targeting and recommendation systems at Quora, URX (acquired by Pinterest) and MIT.