Artificial Intelligence (AI) is poised to have major impacts on many facets of biomedical research in the near future. The recent successes of deep learning in computer vision, natural language processing, game playing and autonomous vehicle development provide hints of what is to come from advances in AI applied to problems in biology and health research. Already major advances have occured through the use of AI methods in drug development, cancer diagnosis, pathology, genome sequence analysis, patient risk analysis and prediction of antibiotic resistance. Today it is becoming routine to build machine learning based predictive models where we have significant amounts of data but lack an underlying theory of mechanism. With careful uncertainty analysis these models can be used in place of traditional mechanistic models in some applications. Generative models can be used to create large collections of synthetic data that models real data, and can be used to generate drug candidates, example DNA sequences and to generate “synthetic patients” for computational analysis. In the future hybrid models that combine traditional first principles mechanistic models with those learned from data will out perform either individual model type.
This session will include speakers developing and using AI accross a range of problems from DNA sequence analysis, to drug design, prediction of phenotypes from genotypes and diagonosis diseases such as cancer, retinal myopathy, lung disease and the interpretion of medical records and to development of treatment strategies.
|Machine learning models of brain ageing in health and disease;
|11:20||Clint Davis-Taylor||Automated Parameter Tuning for Living Heart Human Model using Machine Leaning and Multiscale Simulations|
|11:35||David Wright||Combining molecular simulation and machine learning to INSPIRE improved cancer therapy|
|11:50||Amanda Minnich||Safety, Reproducibility, Performance: Accelerating cancer drug discovery with ML and HPC technologies|
|12:05||Fangfang Xia||Deep Medical Image Analysis with Representation Learning and Neuromorphic Computing|
|12:20||Rick Stevens||Deep Learning in Cancer Drug Response Prediction|
|AI for Big Science
|10:35||Ola Engkvist||Applying Artificial Intelligence in Drug Design|
|10:50||Valeriu Codreanu||The Convergence of HPC and AI for Healthcare on Intel® Based Supercomputers|
|11:05||Justin Wozniak||Accelerating Deep Learning Adoption in Biomedicine With the CANDLE Framework|
|11:20||Gregory Parkes||The Influence of DNA Sequence-Derived Features across the ‘omics scales|
|11:35||Rafael Zamora-Resendiz||Predicting ICU Readmission with Context-Enriched Deep Learning|
|GuacaMol: Benchmarking Models for De Novo Molecular Design|