CancerGPT: AI Can Now Predict the Results of Cancer Treatment Research
AI is craving day by day and you have to be updated about that.
A team of U.S. researchers has developed CancerGPT, an artificial intelligence model that leverages large pre-trained language models (LLMs) to predict the impact of different drug combinations on rare human tissues found in cancer patients. This groundbreaking approach marks a significant leap forward in medical research, especially in areas where structured data and sample size are limited. Brace yourselves for a revolution in the fight against cancer!
Both the University of Texas and the University of Massachusetts have collaborated and harnessed the power of LLMs to extract proir knowledge from a medical research texts and applied it to complex biological inference tasks.
The outcome?
Astounding accuracy! Even with minimal or no samples, CancerGPT achieved significant precision. The team’s experiments, involving seven rare tissues from different cancer types, left no doubt about the model’s extraordinary potential.
2023 has been ablaze with discussions about AI LLMs in medical research. Just recently, Decrypt reported on the creation of Ankh, an LLM that understands how proteins communicate, by a group of visionaries from the universities of Munich and Columbia in collaboration with biotech company Protinea. Furthermore, another research group employed AI technology to identify promising candidates for senolytic drugs, offering hope to slow the aging process and combat age-related diseases by targeting “zombie cells.”
With CancerGPT boasting around 124 million parameters, comparable to the larger fine-tuned GPT-3 model, and utilizing zero-shot GPT-3 for coherent responses, this AI marvel has dazzled the scientific community. While the accuracy of its arguments may occasionally be unverifiable or susceptible to hallucination, the researchers acknowledge the immense value hidden in scientific literature for cancer types with limited structured data.
