Researchers at Columbia University Vagelos College of Physicians and Surgeons have developed a new artificial intelligence method capable of accurately predicting gene activity within any human cell. This breakthrough, published in Nature, could significantly enhance the understanding of diseases such as cancer and genetic disorders.
Raul Rabadan, professor of systems biology and senior author of the study, emphasized the potential of predictive computational models to uncover biological processes swiftly and accurately. Traditional biological research typically describes cellular functions but lacks predictive capabilities regarding cellular responses to changes, such as mutations that cause cancer.
The study utilized gene expression data from over 1.3 million human cells, enabling the AI system to predict gene expression in previously unexamined cell types, aligning closely with experimental results. The AI's capabilities were further demonstrated when it identified mechanisms behind an inherited form of pediatric leukemia, predicting how mutations affect transcription factors critical to leukemic cell fate.
Additionally, the research opens avenues to explore the genome's 'dark matter'—regions that do not encode proteins but may harbor significant mutations. Rabadan's team is investigating various cancers to understand regulatory mechanisms and their implications for developing new treatments.
This innovative approach marks a shift towards a predictive science in biology, potentially transforming research methodologies and therapeutic strategies.