Finding the right medicine to treat conditions like anxiety and depression can be complicated. A doctor will start you in a drug that is usually tolerated and effective, but can do nothing for you-or have terrible side effects. Sometimes it takes months of test and mistake to find something that works.
Next to a very common issue. Dr. Priscilla Chan told an audience south from southwest Wednesday that it could be directed if doctors could control medicines against a generating model of your cells and systems. Chan, who co-founded the initiative Chan Zuckerberg with her husband, the founder of Meta and CEO Mark Zuckerberg, said that use of it could be another great dance for biomedical research.
“Hope is with those models that we will be able to answer some of the most difficult questions in biology,” Chan said.
Artificial intelligence has been a hot topic for almost everyone since his broken moment with chatgpt his chatbot debut at the end of 2022.
Last year, two scientists in the Deepmind’s unit of Google acquired Nobel’s chemistry forms for their work using it to predict the protein structure.
In terms of how this technology can progress science and medicine, it can take years, if not decades. And these models are likely to speed up current laboratory research, not replace it. But Chan sees a world of opportunities.
What do we not know about ourselves
Chan, a pediatrician, said that many of the way the human body works still belongs to the meaning of science. Of course, it has been a few decades since researchers hit the human genome, but Genetics offers only one road map. Chan used the analogy of a Millennium Falcon’s LEGO set of Star Wars – the genetic code is the instructions package. However, we still do not know how individual parts come together to form spaceship. And when part does not seem to fit well, the medicine has to come here.
Beyond gaps in scientific knowledge of biology, we also have a limited understanding of how biology works within individual people. Based on a small number of samples, we have extrapolations of how the body is supposed to function, but this is a small data base that does not approach to represent the full variety of humanity.
A pattern it can help describe what is happening in an individual’s cells – personalizing the drug so that your treatment can change from mine.
“If we build the right data and the models of it, we can better understand what is making us healthy and what is making us sick,” Chan said.
Can he speed up biomedical research?
Current research techniques are also slow and expensive in the development of medicines and new treatments. Ideas should be tested in a physical laboratory environment, which takes a time and extraordinary resources.
Chan does not suggest eliminating existing physical research “wet lab”. But a machinery learning model-a distinctive mark and he-can help identify drug candidates with a higher probability of working, means that it can take fewer real-world tests to achieve a applicable solution.
Models will not always be correct. They will offer non -functioning solutions and ideas, perhaps physically impossible ideas, but that is why there should be a filter of real human scientists dealing with the ideas that a model produces.
“It won’t give us the full answer,” Chan said. “I don’t want you to think that scientists will simply talk with a model and get all the answers they need.”
Machines can help scientists find better questions, Chan said. “It will be the hypothesis generator,” she said.
While many companies and researchers are watching ways to use it in hospitals and patient treatment, Chan’s concentration is to advance basic biological research that makes future advances possible. She sees it as a great jump for science, similar to the invention of the microscope, X -rays, MRI or human genome sequences.
“Health and medicine, it moves in steps,” she said. “It has been decades when the research is stuck, and then someone invents a new technology that completely changes the way we see the human body.”