AI tools are changing the way pharmaceutical companies and public health organizations think about finding drugs to treat rare diseases.
For example, Transcript Bio, a According to a profile in Bloomberg this week, the Silicon Valley-based pharmaceutical startup is using AI to match existing drugs with new use cases.
The company uses AI software to identify drugs that could alleviate the effects of rare diseases. Instead of narrowing down to one condition to see whether an experimental drug works on the genes associated with that condition, Transcripta works backwards by starting with drugs already on the market.
Related: AI is changing the way companies recruit for open positions – and how candidates exploit the system
The startup tests drugs that its AI software has identified as good candidates to determine whether these drugs have additional use cases and also affect genes associated with rare diseases.
Transcripta’s unusual approach could be more cost-effective. The company claims it can screen for all known rare genetic diseases for about $2 million. According to Bloomberg, a pharmaceutical company could pay to begin research on a new drug for a disease.
The company has raised about $30 million so far.
“It’s much more capital efficient now,” Chris Moxham, CEO and chief scientific officer at Transcripta, told the publication. “It’s not just one disease after another.”
Related: A new AI chatbot is revolutionizing business school curriculum and accreditation – it could change that
Other startups, including Century Health, which secured $2 million in pre-seed funding last month, are also trying to leverage existing drug data for new real-time applications. Century Health, in particular, is working with its AI platform to commercialize breakthrough drugs for diseases such as Alzheimer’s.
It’s not just startups that are interested in AI applications in healthcare. AI chip giant Nvidia has developed an AI cloud service specifically for AI-powered drug development and last month released two dozen new AI tools for healthcare.
One tool that stood out was a bot called Hippocratic that outperformed real nurses in everything, including how a drug affects labs and comparing a lab value to a target range.
Related: Microsoft’s new AI can make photos sing and talk
In public health, researchers at the National Cancer Institute (NCI), part of the National Institutes of Health (NIH), last week developed an AI tool called PERCEPTION that showed promise in predicting whether a patient will would target certain types of cancer treatment drugs over others.
In a study published Thursday in Nature Cancer, researchers wrote that PERCEPTION outperformed the industry standard for predictors “in all clinical cohorts.”