Jose-Luis Ambite, PhD

Director of Data Science

SC CTSI Role

Contact Information

ambite@isi.edu
310-448-8472

Professional Background

I have research interests in data integration (including schema mapping, record linkage, efficient query rewriting, and information extraction), databases, knowledge representation, semantic web, machine learning, federated learning, genetics and biomedical data science. In the last 15 years, my research focus has been on developing novel approaches for integration, analysis, and dissemination of biomedical and genetic data. Some research and project highlights: * Data Integration. I chaired the Information Integration Working Group of the NIGMS-funded Biomedical Informatics Research Network (BIRN), a large multi-site project to improve the infrastructure for data sharing in biomedical research. I led the development of the BIRN mediator, a virtual data integration system which was applied in a variety of domains, ranging from functional MRI studies, to non-human primate genetics and pathology data, to cardio-vascular disease, and to radiation oncology. I designed scalable algorithms for query rewriting, a core problem in data integration, which were two orders or magnitude faster than the previous state-of-the-art. I also was a core designer of the automatic schema mapping generation algorithms of the Karma data integration system. In the NIMH-funded SchizConnect project, we extended the BIRN mediator to integrate several leading repositories of clinical, imaging and genetic data on Schizophrenia. I was Principal Investigator of the Data and Software Coordination and Integration Center of the NIBIB-funded Pediatric Research using Integrated Sensor Monitoring Systems (PRISMS) Program, where we developed a scalable architecture for integration and analysis of sensor, environmental, and clinical data to better understand pediatric asthma. * Genetics. I led the data integration and data dissemination activities of the Coordinating Center for the NHGRI-funded Population Architecture using Genomics and Epidemiology (PAGE) network, a consortium of four major studies (CALiCo, MEC, WHI, ISMMS) with the goal of understanding the association of genetic variants with complex diseases and traits across populations. Within this context we developed iLASH, currently the most efficient and accurate identity-by-descent detection algorithm. I also lead the data integration, automatic data curation, and dissemination activities of the NIMH Repository and Genomics Resource (NRGR), which allows researchers to query across the phenotype and genotype to identify relevant individuals with mental disorders or controls, and their biosamples. * Machine Learning and Federated Learning. I was Principal Investigator of the DARPA-funded Secure Heterogeneous Learning Federation with Information-Theoretic Guarantees (SHELFI), which developed federated learning methods that allow training deep learning models without sharing data and with strong privacy guarantees. Finally, I am mPI of the NIA-funded “Federated Deep Learning to Accelerate Alzheimer's Disease Research” project that establishes a global study with sites in India, Japan, Spain, and the US to study Alzheimer’s disease powered by federated learning I received an Electrical Engineering degree from the Technical University of Madrid (UPM-ETSIT) and a PhD in Computer Science from the University of Southern California. I received a Fulbright / Ministry of Education and Science of Spain Scholarship.

Interests

Outside of work, Dr. Ambite enjoys travel and martial arts.