Dozens of Seattle Biotech companies are using artificial intelligence for new treatment treatment design. However, at a conference of industrial leaders and investors this week, scientists have given a brief message: AI has a lot of promises, but the expectations must be based on reality.
Trade association Life Science Washington And investment firm Madron Dealing at Biotech, Pharma and AI, he hosted the ODI forum in Seattle in the suburbs.
“Some of these AI models have some over -hype on the depth and depth of affordability,” said Jamie LazarvitsCEO and co-founder Archana BiosciencesThe He said that researchers need to “really be careful” to draw decisions from the data produced by models.
And still be excited about a lot of things.
“What was the fiction of science 15 years ago is reality now,” Eric Proco, the chief scientist said Cyrus biotechnologyThe “So yes, Hyp is but there is still a lot of potential and sometimes the progress that is being made is simply relaxing” “
Lazarovits and Proco were part of a panel with four Seattle Startups AIK. Each company is dealing with several challenges in the development of drugs:
- ArrowA company that originally originated from stealth with Ste 20 million financing, to design antibodies or ABC -owned protein structures using AI, which is intended to help antibodies target cells and avoid other cells.
- Cyrus The design of the immune system is creating a challenge by focusing on the detection and removal of the regions that trigger the reaction of the immune-system. The 10 -year -old company has collected the .6 36.6 million dollars according to the pitchbook.
- Outpace bioA startup established in 2021, which has collected $ 200 million, is an engineering protein that aims to encourage T-cell therapy for solid tumors, which form 90% of cancer. Tumors are discouraged in removing current T-cell treatment, which often stops working within a month.
- Talaxi BioscienceWhich was launched in 2021 and has about $ 20 million, the transcript factor aims – proteins that are “control” parts that turn off the genes and turn on. The company is targeting transcription factors that activate genes that run certain cancer.
Beyond discussing their own work, panel members identified the basic principles of how to do – and how to do in biotech study – how to do:
Researchers do not replace, their
Mark LazoOutpace’s co-founder and CEO, compared AI equipment to the Robotic Exoscaleton donated by others in Science-Fi Film Aliens to remove heavy cargo-and fought against the At Genomorf Quin.
“It makes the researcher better,” he said. We’re not trying to replace the researcher.”
Models must meet the reality
Lazarovites noted that although AI can generate exciting lead and information, it does not mean too much until the actual test with cells and organisms.
“Whenever we try to adopt the new model, new AI methods, you can get incredibly excited to get this silicon validity,” he said. “But the reality is, what are you actually validate in the wet lab?”
Real barrier: clinical trial
AI engineering is great for new therapy, but the most expensive, laborious part of the drug development process is seeing how they work for patients.
“The most effective place to really change the game for AI development for AI is to create a smaller, better -driven clinical study,” said Lazo. The way to do this, he added, that the performance of multiple tasks was bringing better drug candidates.
Still searching for the epoch -like moment of AI
Proco is still awaiting AI to move forward in the advancement of encouraging biotech and pharmacology.
“AI is great nowadays to predict the protein structure, but it has not yet found its killer app to make new drugs,” Proco said, “Proco said. “AI is letting us make new drugs now that was impossible to make before? How is it being a game changer?”
This question has captured a central tension discussed at the panel and conference: though AI Seattle’s biotech companies have converted to drug design, the industry is still navigating the gap between calculating and clinical evidence.