12:49 PM
 | 
Jun 07, 2019
 |  BioCentury  |  Emerging Company Profile

insitro: machine learning with designated data

How insitro aims to improve machine learning-based target discovery by creating its own ‘fit-for-purpose’ data

Just 12 months after announcing the launch of insitro, AI veteran and CEO Daphne Koller has led the drug discovery and machine learning company to a series A round of over $100 million and a biopharma deal with Gilead. insitro’s core selling proposition is that it generates "fit-for-purpose" data that takes much of the noise out of machine learning-based target discovery.

Most companies using machine learning to discover targets aggregate previously collected data sets to feed into their software. These can be from genomic, molecular or clinical studies.

However, Koller said the approach can lead to selection bias, batch effects and other artifacts because the data were not generated for the purpose of the AI analyses. It’s also not possible to conduct new experiments on these data sets to evaluate, for example, what would happen if a gene was altered.

"It's difficult to get the right value from them," she said. “We have the willingness and ability to create data sets that fit the purpose of machine learning.”

insitro Inc....

Read the full 830 word article

User Sign in

Trial Subscription

Get a 4-week free trial subscription to BioCentury

Article Purchase

$150 USD
More Info >