London-based startup Basecamp Research has closed a $60 million Series B round that will help expand its genomic database and what it sees as gaps in biological information that will be needed to build more powerful AI models.
Also on Wednesday, the startup announced that it has signed a multiyear deal with David Liu’s laboratory at the Broad Institute of MIT and Harvard. The two groups will work on programmable genetic medicines but are keeping further financial and scientific specifics under wraps.
Basecamp is looking to fill the data gaps that prevent the training of more powerful artificial intelligence models in biology. Models are only as good as the data they are trained on, and many areas of biology lack the quality datasets needed to build better models.
The company’s solution is to visit the most exotic places in the world and mine hard-to-get genetic data. The company was founded in late 2019, shortly after its two co-founders, CEO Glen Gowers and Oliver Vince, completed a three-person expedition to the Vatnajökull glacier in Iceland. Capturing samples of air, soil and water, while looking for organisms large and small, they collected little-studied species.
“The majority of our data was completely unannotatable or unrecognizable relative to existing reference databases,” Gowers said in an interview.
Gowers compared the vast number of species on Earth to the Atlantic Ocean. Scientists have described the genomes of species for what would be the equivalent of five drops of water. “That, for us, felt like a giant gap,” he said.
The Series B was led by the European VC firm Singular. Gowers said it took four weeks to complete the round, which initially aimed to raise $45 million. Other investors include S32, True Ventures, Hummingbird Ventures and André Hoffmann, the vice chairman of Roche, among others. Since its founding, Basecamp has now raised $85 million.
The biotech has also hired industry veteran Anupama Hoey as its chief commercial officer, who will look to sign more biopharma deals that tap into its growing genetic database.
From deep-sea dives to building better AI
Basecamp’s 34 employees include not only data scientists and machine-learning engineers, but also deep-ocean divers and professional explorers. They have completed over 100 expeditions in over 20 countries, from Costa Rica to Cameroon, to add more species to their database. This past weekend, a team set off to visit a volcano in the middle of the Atlantic Ocean.
Basecamp forecasts that its database will grow 10 times in size by year’s end and jump another order of magnitude within a couple of years. Already, its leaders say its dataset is 10 times larger than widely used, publicly available comparators. It expects its dataset will be 1,000 times larger than what’s publicly accessible today in about two years.
“We, as a society, do not know enough about nature,” Vince said. “Basecamp is in a position to go out and solve it.”
Basecamp has also shifted away from “helicopter science,” as Gowers called it, and now favors signing deals with developing countries and working with local scientists. That reflects longstanding ethical concerns over how to work in biodiverse parts of the globe, particularly under agreements like the Nagoya Protocol, which tries to share the benefits of genetic resources with those nations.
There’s also a question of how useful the data will be for drug R&D. Last year, Basecamp presented research showing its database could boost the performance of AlphaFold2 in predicting protein structures.
The Liu lab collaboration may shed some additional light on its utility. The two groups will work on novel fusion proteins, which are used in gene-editing technologies like base and prime editing. Specifics are limited on the collaboration, but Gowers said potential benefits could be proteins that are smaller, more efficient, and easier to manufacture.
Beyond the Liu deal, Basecamp has about 15 commercial partnerships across various industries, including with a few undisclosed pharma companies.
Basecamp’s 30-year-old co-founders believe its data can help unravel the complicated web of relationships among biomolecules. Vince compared it to how Facebook aggregates data to better understand a person’s role in society. Basecamp’s goal is to eventually do the same, but for proteins and other biomolecules.
“The same approach can be applied to biology if you have enough context,” Vince said.