Priscilla Chan Sees AI Models of Cells as the Next Leap in Biology and Medicine
Finding the right medicine to treat diseases like anxiety and depression can be tricky. Doctors will start using a drug that is usually well tolerated and effective, but it can do nothing to you, or has horrible side effects. Sometimes it takes months of trial and error to find something that works.
This is a very common problem. Dr. Priscilla Chan said Wednesday at South The South audience in the Southwest that doctors can simplify it if they can check the drug based on the AI model of your cells and your system. Chen, he co-founded Chan Zuckerberg Initiative Meta founder and CEO Mark Zuckerberg, along with her husband, said using AI could be the next big leap in biomedical research.
“Hopefully with these models we can answer the most difficult questions in biology,” Chen said.
Artificial intelligence has been a hot topic for everyone since Chatgpt’s debut Broadcast Second half of 2022. This week, this is the main focus of the Austin, Texas SXSW believe,,,,, Accountability and The future of work.
Last year, two scientists in Google DeepMind AI unit Win the Nobel Prize Chemically use AI to predict the structure of proteins.
As for how the technology advances science and medicine, it may take years, if not decades. These AI models may only speed up actual laboratory research, rather than replace it. But Chen saw a world of possibility.
We don’t know about ourselves
Pediatrician Chen said the way the human body works is still incredible. Of course, it has been decades since researchers cracked the human genome, but genetics only provides a roadmap. Chan used an analogy of the Millennium Falcon’s LEGO kit from Star Wars – the genetic code is the instruction pack. However, we still don’t know how the debris merged together to form a spacecraft. And, when a part seems inappropriate, this is where medicine needs to be involved.
In addition to the gap in scientific knowledge about biology, we also have limited ways in which biology works within individual characters. Based on a few samples, we infer how the body should work, but this is a tiny dataset that does not represent the pure diversity of human beings.
AI models can help describe what is happening in a person’s cells – personalized medicine that makes your treatment different from mine.
“If we build the right data and AI models, we can better understand what makes us healthy and what makes us sick,” Chen said.
Can AI speed up biomedical research?
Current research techniques are also slow and expensive in developing new drugs and treatments. Ideas must be tested in a physics laboratory environment, which requires a lot of time and resources.
Chen does not recommend eliminating existing physical “wet lab” studies. However, machine learning models (the logo of AI) can help identify drug candidates with higher probability of working, meaning that implementing a viable solution may require fewer real-world testing.
These models are not always correct. They will provide solutions and ideas that cannot be solved, perhaps physically impossible ideas, but that is why there is a real human scientist to solve the ideas produced by the model.
“This won’t give us all the answers,” Chen said. “I don’t want you to think scientists will just talk to models and get all the answers they need.”
Chen said these machines could help scientists find better problems. “This will be a hypothesis generator,” she said.
While many companies and researchers are looking for ways to use AI in hospitals and treatments for patients, Chan’s focus is on advancing fundamental biological research that makes future advances possible. She believes that AI is a potential major leap in science, similar to the invention of microscopy, X-ray, MRI or sequencing of the human genome.
“Health and medicine, it’s going leaps and bounds,” she said. “For decades, research was stuck and then someone invented a new technology that completely changed the way we see the human body.”
