Monoclonal antibodies have transformed the treatment of many diseases from cancer to autoimmune disorders. But the approach has its limitations, prompting scientists and companies to develop the next generation of antibody treatments and other biologics.
“We appreciated this is a great way to go for developing new therapies, but at the same time, we realized that creating these next-generation antibodies is much, much more challenging than creating monoclonal antibodies,” said Mark DePristo, Ph.D., co-founder and CEO of BigHat Biosciences, a company developing an artificial-intelligence-based antibody engineering platform.
The company raised a $19 million series A round from Andreessen Horowitz, 8VC, AME Cloud Ventures and Innovation Endeavors to further develop its platform, make it scalable and advance a suite of programs toward the clinic. It will also use the funds to build the teams that will make this possible, with plans to hire protein engineers, artificial intelligence and machine learning engineers, and computational biologists.
At its core, BigHat’s technology is designed to reduce the time it takes to get from an antibody design—a string of amino acid residues that comprise a protein—to creating that molecule in a lab from weeks or months to just days, Peyton Greenside, Ph.D., chief scientific officer of BigHat, said.
“Traditional approaches focus on bulk screening, looking at a large population of designs and trying to pull out something of interest according to one property,” Greenside said. That property could be affinity, or the antibody’s ability to bind to its target, or stability, so the medicine can be stored or transported without special requirements.
What sets BigHat’s tech apart is its ability to measure all of these properties at once for a given candidate rather than starting with one and hoping the others will follow.
“Every time you update your model with data that comes off the platform, you can use what you learned to redesign molecules with improved properties you care about,” she added.
“BigHat brings together synthetic biology, machine learning, and automation to offer a general, reliable, fast and flexible platform for engineering antibody therapeutics,” Vineeta Agarwala, a general partner at Andreessen Horowitz who joined BigHat’s board, said in a statement. “Their integrated approach solves challenges in protein engineering across many therapeutic areas. AI is modeling the molecular properties of sequences, supercharging the process to eliminate challenges currently slowing down drug development.”
Over the next few years, BigHat plans to scale the capacity of its platform up to 100-fold so it can pursue its own programs as well as programs with partners, DePristo said.
Those partnerships could include working with companies that have already developed antibodies but are struggling with certain properties, like stability or manufacturing.
“We’d like to work with those companies to improve the antibody,” DePristo said. “We have existing partnerships today and we’re looking to expand those. We see this as a key way for BigHat to get our platform technology into the community to produce better molecules as quickly as possible.”
Another avenue for partnerships is to enlist partners in developing completely new antibodies that come out of BigHat’s platform.
“We see those opportunities in front of us and are looking to pursue them, in the near term, internally in BigHat,” DePristo said. “Whether we partner them out in the long term or develop them end-to-end is currently TBD.”