HKU team creates AI tools to boost cancer diagnosis

HKU research team led by associate professor Luo Ruibang developed two AI tools to improve cancer diagnosis and RNA research.

The algorithms, ClairS-TO and Clair3-RNA, use long-read sequencing technology to detect gene mutations in complex samples.

Standard cancer diagnosis usually requires comparing tumor samples with healthy tissue, but healthy samples are not always available. HKU says ClairS-TO analyzes tumor DNA without paired samples by using a dual-network system to confirm mutations and remove errors.

Clair3-RNA uses deep learning to distinguish real mutations from biological noise and RNA editing. This allows clinicians to analyze gene expression and mutations at the same time with high accuracy.

The breakthrough will make genetic analysis more accessible and support personalized medicine. The research was published in the journal Nature Communications.