Event

Photo: Wei Qui

Abstract: In the midst of ongoing challenges such as aging and disease management, which are as multifaceted as they are critical, lies a common computational challenge—how to effectively harness Artificial Intelligence (AI) to predict and explain complex biological phenomena. This talk will explore novel AI techniques that provide both predictability and transparency across various aspects of biomedicine by integrating Explainable AI (XAI) into biological research.

First, I will introduce my work in AI for Personalized Health Insights. I will present the ENABL Age framework, which integrates AI and XAI to provide precise and interpretable assessments of biological aging. This model not only estimates biological age but also explicates the factors contributing to aging, offering insights for personalized health strategies.

Second, I will discuss my AI innovations in Omics Data Analysis. I have designed DeepProfile, which analyzes large-scale cancer datasets to identify key biomarkers and pathways, enhancing our approach to precision oncology. Additionally, I have developed StrastiveVI, which further isolates aging-related signals from single-cell transcriptomic data, revealing universal aging patterns and facilitating targeted anti-aging interventions.

Third, the discussion will turn to our pioneering work in the automated generation of lay language summaries, which makes complex biomedical findings more accessible and boosts public health knowledge. This work improves biomedical communication by making health-related information more comprehensible to the public.

In conclusion, I will outline a vision for future directions in bridging transparent AI with biology. This effort demands collaboration across biology, clinical science, AI research, public health, and data science. Such a multidisciplinary approach will tackle current challenges in understanding complex biological processes and pave the way for innovative solutions that improve both the precision and accessibility of healthcare.

Wei Qiu website.

Twitter/X: @weiqiu55