Research

Collaborative AI research to advance biomedical frontiers from disease diagnosis and drug discovery to single-cell multi-omics, with a commitment to precision, reproducibility, interpretability, and trust.

Personalise

Theme 1: AI-Enhanced Disease Diagnostics and Personalized Medicine

New sequencing technologies enable the comprehensive analysis of various molecules (DNA, RNA, proteins) efficiently and affordably, generating vast amounts of omics data (genomics, transcriptomics, proteomics) from patient samples or individual cells. This data, alongside wearable sensor readings, electronic health records, and medical imaging, offers significant potential to advance personalized medicine and precision therapy. However, realizing this potential requires overcoming challenges through sophisticated AI models and big data analytics for effective data interpretation, prediction accuracy, and decision reliability. At the Biomedical AI Laboratory, we focus on improving disease diagnostics with generalizable models, enhancing non-invasive cancer management via better liquid biopsy tests, and increasing test sensitivity by incorporating multi-omics data and other patient information.

Industry partners: , , , (CSIRO/Data60 NGGP)

Theme 2: Single-Cell Multi-Omics: Tackling the Big Data Challenge

The advent of single-cell multi-modal omics (scMulti-omics) technologies has revolutionized our ability to measure diverse molecular features—like DNA methylation, chromatin accessibility, RNA expression, and protein levels—in individual cells, enabling a comprehensive understanding of cellular functions. Recognized by Nature Methods as the 2019 Method of the Year for its groundbreaking insights into cell functionality, novel cell-type discovery, and cross-omics relationships, the scMulti-omics field has rapidly grown. Despite this technological progress, the field faces significant computational challenges in integrating and interpreting the vast, complex datasets generated, hindering the accurate prediction of biological phenomena. Issues such as data heterogeneity, noise, systematic biases, and the curse of dimensionality are prevalent. At the Biomedical AI Laboratory, we employ AI to navigate these obstacles, enhancing the utility of single-cell data in life science research and clinical application.

Industry partners: ,

Theme 3: AI-Enabled Drug Discovery

AI is transforming drug discovery and development, offering a revolutionary approach to the pharmaceutical industry. By analyzing complex datasets, including clinical, biological, molecular, and genetic information, AI algorithms enhance the precision and efficiency of drug development from drug target identification to toxicity assessments and compound design optimizations. Accordingly, AI facilitates the streamlining of research processes, from preclinical to clinical study designs, and supports the repurposing of drugs for new therapeutic applications.

We leverage advanced AI techniques to mine and interpret vast molecular, structural, and clinical datasets. This aids in repurposing drugs, identifying effective drug combinations, elucidating drug targets and mechanisms, and forecasting drug interactions and side effects, streamlining the path to novel therapeutics.

Industry partners: , (CSIRO/Data60 NGGP)