Alphabet’s Verily has developed a new machine learning approach for mass spectrometry that can interpret data and identify new disease biomarkers. Aptly named DeepMass, the method has shown to be more accurate in early tests than an existing predictive model and also enlarged the scope of known biomarkers when applied to clinical data.
The life sciences company evaluates protein profiles using mass spectrometry as part of their work on several projects – including The Project Baseline Health Study – to learn more about how diseases develop in the body.
The conventional method of developing spectral libraries can be both labor and time-intensive. With this in mind, Verily was curious about whether there was a better way. They partnered with the Max Planck Institute of Biochemistry and Google to look at whether machine learning could play a role in the process.
The final product is an approach that can create spectral libraries through computation. Verily’s thought was that they could more quickly generate the reference materials to use for decoding large datasets of protein markers from mass spectrometry and the first results are promising.
According to Verily’s blog article on the subject, researchers discovered that not only are “DeepMass-calculated spectral libraries… equivalent to the experimental ones”; they also learned that “the model correctly learned known chemical rules that govern a peptide fragmentation, but also suggested some new ones.”
Verily has chosen to make their service available using Google Cloud Machine Learning Engine. The hope is that by sharing the tool, it can help developers of diagnostics differentiate disease-related protein profiles.
“We hope that DeepMass available on Google Could, will enable researchers to characterize disease-relevant protein profiles to build new diagnostic tools and therapeutics,” Verily researchers Peter Cimermancic and Roie Levy wrote.
Verily’s machine-learning tool is just the latest innovation offered through their ever-expanding research and development business. They also have a blood-glucose monitoring partnership with Dexcom and a robot surgery joint undertaking with Johnson & Johnson.
The advent of artificial intelligence has made a significant impact on healthcare and Alphabet’s Verily is a leader in that industry. It’s a natural evolution for Google – whose business began it all by building algorithms that moved modern life through the web.
“The fundamental underlying technologies of machine learning and artificial intelligence are applicable to all manner of tasks,” said Greg Corrado, a Google neuroscientist. He believes that’s true “whether those are tasks in your daily life, like getting directions or sorting through email, or the kinds of tasks that doctors, nurses, clinicians and patients face every day.”
Healthcare is a $3.4 trillion industry and Verily’s development of research services and solutions will prove to be lucrative in the future. “It’s pretty hard to ignore a market that represents about 20 percent of GDP,” said John Moore, an industry analyst with Chilmark Research.
