Killing Diseases with Really Big Computers: Building Analysis Tools to Solve Disease| Marisa Torres
About This Video
Marisa Torres, Bioinformatics Lead, Lawrence Livermore National Lab (LLNL) presents the Technical Vision Talk “Killing Diseases with Really Big Computers: Building Analysis Tools to Solve Disease”. Our team at LLNL has been working on improving bioinformatics tools and models for COVID-19 and for cancer. We think rapid response to a disease outbreak should be a national security priority. We’ve run huge gene simulations for drug discovery and built machine learning models on the results. For COVID-19, we’ve aimed to design viral inhibitors with no adverse health reactions, and we’ve successfully released most of this work publicly as a searchable and usable tool. We’re now scaling up our work from a few COVID-19 genes, up to tens of thousands of human genes for the American Heart Association. When designing therapeutics, we use every informatic and statistical tool available. We create new machine learning strategies for scaling virtual screens for the human or microbe. We use 3D models and docking poses of protein structures to make predictions. We screen for safety properties, because we don’t want detrimental interactions.
In This Video
Bioinformatics Group Leader, Lawrence Livermore National Lab
Marisa designs, implements, and integrates data in relational databases, provides software engineering support for DNA signature discovery, and responds to internal and external customer requests for signature analysis. She has provided signature development and bioinformatics analysis for the Environmental Protection Agency and the National Bio-Forensics Analysis Center. In 2000, Marisa designed DNA signatures that were promoted for use in the BASIS program. She has taken the lead on signature erosion checking, which during the most recent DHS proposal cycle was recognized as important for continued reliable detection of pathogens, and she supports public health and biosecurity customers combining her versatile skill set of software engineering and biology background.