Guillermo del Angel

Exec. Dr., Rare Disease TA Head, Centre for Genomics Research
AstraZeneca

Guillermo del Angel, PhD is currently the Rare Disease Therapy Area Head as well as the head of the Data Science team at AstraZeneca’s Centre for Genomics Research in Boston, MA, USA. There, he has been leading efforts to use human genetics and multi-omics to improve rare disease target discovery, drug development and improve our disease understanding. He is also leading efforts to use state of the art Machine Learning and AI approaches to improve the generation of biological insights from multi-omic data and to improve disease prediction and patient care.
Before this role, he held several positions at Alexion Pharmaceuticals (later Alexion, AstraZeneca Rare Disease) since joining in 2015, including serving as Alexion’s Head of Bioinformatics and Data Science. He was also formerly a consultant at McKinsey & Co. and a computational biologist in the Medical and Population Genetics program at the Broad Institute of MIT and Harvard.
A native of Mexico City, he holds a BS degree from ITESM in Mexico, a MS from Boston University and a PhD from Cornell University. 

 

AI-Driven Multi-Omics and Biobank Partnerships: Accelerating Innovation in Drug Discovery

Genomics offers massive opportunities for improving drug development success rates. Currently, 90% of pipeline drugs fail, but targets with genetic support are 2x more likely to gain approval, and 66% of FDA-approved drugs have human genetics evidence supporting them. I will show in this talk how AstraZeneca’s Centre for Genomics Research is empowering AstraZeneca’s pipeline through novel uses of genomic and multi-omic technologies, anchored by novel AI platforms and leveraging global biobank partnerships.
 
AstraZeneca’s Genomics Initiative was launched back in 2016 with the goal to analyze up to 2 million genomes, integrating genomic, proteomic, and clinical data to understand disease biology, identify new drug targets, and support precision medicine strategies. I will first describe how analysis of over 480 thousand UK Biobank exomes revealed thousands of significant gene-level associations, with rare variants showing significant effect sizes and providing direct biological insights. The MAP3K15 diabetes research exemplifies this approach—rare variants associated with reduced diabetes risk showed complete loss of function provided greatest protection, with safety analysis revealing no adverse associations, providing a compelling novel target hypothesis for further therapeutic development.
 
I will then how increasing genetic diversity of the biobanks and expanding our insights across different ancestries empowers drug discovery. I describe how our work with Genes & Health, a community-based genetic initiative aimed at improving the health of Pakistani and Bangladeshi ancestry people, has yielded novel scientific insights by analyzing around 55,000 exomes.
 
I will finally describe how combining protein and genomic data enhances drug discovery by identifying protein quantitative trait loci (pQTLs), with exome sequencing uncovering thousands of rare genotype-protein associations, with ~80% not detected in standard genome-wide association studies. This approach provides crucial insights for biomarker discovery, safety profiling, and understanding disease mechanisms. I will then describe how our MILTON AI system predicts over 1,000 diseases, combining clinical biomarkers and 3,000 plasma proteins. MILTON significantly outperformed traditional polygenic risk scores across multiple diseases, with proteomic data substantially improving prediction performance for conditions like multiple myeloma, psoriasis, and Alzheimer’s disease.