Agenda

Health Care is undergoing transformation globally. The need for High Performance Computing and AI Infrastructure to Scale Genomics Sequencing, analysis and artificial intelligence is critical to drive faster discoveries and more effective research. In this session, we will review Goals and Challenges in Health Life Science for High Performance Computing and what Intel, the Broad Institute and the ecosystem partners are doing to contribute to this highly focused segment. We will also review a novel approach in design and introduce new opensource tools to help accelerate Artificial Intelligence inferencing in Genomics. We will take a deeper look at the latest enhancements for Genomics Analysis Tool Kit (GATK) running on Intel’s latest 3rd Generation Scalable processor and the software optimizations enabling revolutionary performance. We will conclude with a scalable way to deploy with Intel Select Solutions for Genomics.

The development of next-generation sequencing (NGS) technologies has led to a significant reduction in the cost of DNA sequencing. The advent of NGS, coupled with data analytics, AI, and advancements in HPC storage and computing technologies, is providing significant opportunities to identify actionable insights from genomics data. The affordability of NGS methods and the increased computing and storage capacities are allowing genomics to be performed at the population level. This has led to the emergence of various large national genomics initiatives worldwide, which aim to capture the genetic variation of their populations to eventuate precision medicine.
Scaling up input data from exomes and whole genomes, scaling up production, and the corresponding challenges created for HPC infrastructure are the key areas of concern in population-level genomics efforts. Currently, 75% of analysis stages in genomics take place in an HPC environment. Hence, scaling out genomics production largely depends on the HPC technologies made available for genomics applications.

In this talk, I will disclose how we have been building up the gene-related information processing, life-style recorder, open environmental database and EMR mining infrastructure for precision medicine gradually since 2016. Precision medicine, an emerging approach for disease prevention, prediction and treatment that takes into account people’s individual variations in genes, environment, and lifestyle data. We have built a gene-disease mutation viewer tool, called Mviewer, which can analyze the single nucleotide variants (SNVs) data with various annotations from different genome databases, as our preprocessing tools to decrease the numbers of candidate SNVs. Then, in this research, we describe a novel tool (AI Variants Prioritizer) based on GeneReviews, Online Mendelian Inheritance in Man (OMIM) and biomedical literatures (PubMed) to rank SNVs from Mviewer as well as the phenotype keywords automatically extracted from electronic medical records (EMRs) to get the most viable disease-causing candidates. On the EMR part, we have developed an auto ICD-10 auto-coder to transform the free text medical records to ICD-10 codes for the reimbursement from the health insurance program. On the life style part, we have developed an App that can be found in Apple Store and Google Play, called 台大醫神 (NTU Medical God), and a backend platform to record or infer the life style of the user which including, exercise, eat, drink behavior, vacation, religion, political party, sleep quality, walk distance, etc. On the environment part, we collect the open data from the Environment Protection Administration, Central Weather Bureau, and Water Resource Administration, Taiwan, as well as the airbox installed in the user’s home to get the PM2.5, Temperature, Humidity data.

Artificial intelligence (AI) is transforming medical and healthcare and, in fact, it is bringing a paradigm changes. Several aspects will be expressed in this talk. Representative AI applications in medical and healthcare will be presented, including but not limited to medical imaging, mental healthcare, clinical decision support, and drug development/drug discovery. In addition, in response to the current pandemic, the talk will touch on topic regarding how AI can assist to combat COVID-19 or even prepare for future epidemic outbreaks. From the perspectives of a government-sponsored AI research center with some personal points of view, All Vista Healthcare Center (MAHC), one of the two sub-centers of the MOST Joint Research Center for AI Technology and All Vista Healthcare at National Taiwan University (AINTU), will be briefly introduced and research projects selected to demonstrate potentials to highly impact medical and healthcare will be showcased. Finally, challenges and opportunities for AI in improving human health will be discussed. At the end of the day, collaborations to form public-private partnerships for landing AI solutions play a key role for future medical and healthcare AI development.

Will be updated soon

Advances in high-throughput biotechnologies have given rise to the increasing of the massive amounts of omics data, such as genomics, proteomics, and transcriptomics. Network Medicine is an application of network biology that provides the quantitative platform for integrating big omics data to capture the impact of perturbations during the disease progression on these complex intracellular processes. Constructing proper network models is a challenging problem that heavily depends on the type and size of the data, especially for future rapid accumulation of personalized omics data. High-Performance Computing (HPC) and cloud computing systems offer opportunities to instantaneously generate and access personalized networks for the revolution of big omics data and translation to precision medicine. This session demonstrated performance comparisons and usage examples in identifying potential biomarkers and designing therapeutic interventions. The combination of these technologies will enable the development of next-generation personalized medicine and clinical decision support systems.

Will be updated soon