Scientists from the Australian National University, the National Cancer Institute of the United States, and Pangea Biomed have successfully developed an artificial intelligence (AI) model called DeepPT, which can help doctors choose the best therapy for cancer patients. The relevant research paper was published in the latest issue of the journal Nature Cancer. DeepPT works by predicting the messenger RNA (mRNA) profile of patients. MRNA is crucial for generating proteins and is also a key molecular information for personalized cancer drugs. Researchers say that when combined with another tool called ENLIGHT, 'DeepPT' can predict patients' responses to various types of cancer therapies. Choosing the appropriate therapy for cancer patients is very important. Researchers trained DeepPT using data from over 5500 patients with 16 common cancers, and the results showed that the treatment efficacy increased from 33.3% without using the model to 46.5% after using it. The researchers stated that 'DeepPT' is based on an AI tool they previously developed for classifying brain tumors. Both AI tools have created microscopic images of patient tissues - histopathological images. These images reduce the time delay in processing complex molecular data. (New Society)