The model, called CODE-AE, has been developed under the direction of Lei Xie, professor of computer science, Biology and Biochemistry at the CUNY Graduate Center and Hunter College. It was able to theoretically identify personalized medicines for more than 9,000 patients, with which their diseases could be better treated than has been the case to date.

Accurate and robust prediction of patient-specific responses to a new chemical compound is critical for the discovery of safe and effective therapeutics and for the selection of an existing drug for a particular patient.

According to the researchers, cell cultures and tissue models are not sufficient to assess efficacy, and disease models can only be transferred to the human situation to a limited extent, thus machine learning can help solve the problem. This would also help reduce the time and cost of development.
 
The model has been presented in the journal Nature Machine Intelligence:
He, D., Liu, Q., Wu, Y. et al. A context-aware deconfounding autoencoder for robust prediction of personalized clinical drug response from cell-line compound screening. Nat Mach Intell 4, 879-892 (2022). https://doi.org/10.1038/s42256-022-00541-0

Source:
https://www.gc.cuny.edu/news/new-ai-model-can-accurately-predict-human-response-novel-drug-compounds