Last year, Salesforce, the company best known for its cloud sales support software (and Slack), pioneered a project called ProGen to design proteins using genetic artificial intelligence. A research moonshot, ProGen could — if brought to market — help uncover medical treatments more cost-effectively than traditional methods, the researchers behind it he claimed in a January 2023 blog post.
ProGen culminated in research published in the journal Nature Biotech showing that artificial intelligence could successfully generate the 3D structures of artificial proteins. But, beyond paper, the project wasn’t much at Salesforce or anywhere else — at least not in the commercial sense.
Until recently that is.
One of the researchers responsible for ProGen, Ali Madani, created a company, Rich, which hopes to bring similar protein-making technology out of the lab and into the hands of pharmaceutical companies. In an interview with TechCrunch, Madani describes Profluent’s mission as “inverting the drug development paradigm,” starting with patient and therapeutic needs and working backwards to create a “tailored” treatment solution.
“Many drugs—enzymes and antibodies, for example—are made of proteins,” Madani said. “So this is for patients who would receive an AI-designed protein as a drug.”
While in Salesforce’s research department, Madani is drawn to the parallels between natural language (eg English) and the “language” of proteins. Proteins—chains of linked amino acids that the body uses for a variety of purposes, from making hormones to repairing bone and muscle tissue—can be treated like words in a paragraph, Madani discovered. Data about proteins fed into a generative AI model can be used to predict entirely new proteins with new functions.
With Profluent, Madani and co-founder Alexander Meeske, an assistant professor of microbiology at the University of Washington, aim to take the idea a step further by applying it to gene editing.
“Many genetic diseases cannot be corrected [proteins or enzymes] lifted directly from nature,” Madani said. “Furthermore, gene editing systems that are mixed and matched for new capabilities suffer from functional trade-offs that severely limit their reach. Instead, Profluent can optimize multiple features simultaneously to achieve a custom design [gene] editor that perfectly suits each patient.”
It’s not out of left field. Other companies and research groups have shown viable ways that genetic AI can be used to predict proteins.
Nvidia in 2022 released a genetic AI model, MegaMolBART, trained on a dataset of millions of molecules to search for potential drug targets and predict chemical reactions. After trained a model called ESM-2 on protein sequences, an approach the company claimed allowed it to predict the sequences for more than 600 million proteins in just two weeks. And DeepMind, Google’s artificial intelligence research lab, has a system called AlphaFold that predicts complete protein structures, achieving speed and accuracy that far exceeds older, less complex algorithmic methods.
Profluent trains artificial intelligence models on massive datasets—datasets with more than 40 billion protein sequences—to create new as well as improve existing gene editing and protein production systems. Instead of developing treatments itself, the startup plans to work with outside partners to produce “gene medicines” with the most promising pathways to approval.
Madani claims this approach could dramatically reduce the time — and capital — typically required to develop a treatment. According to industry group PhRMA, it takes an average of 10-15 years to develop a new drug from initial discovery through regulatory approval. Recently calculates Meanwhile, the cost of developing a new drug ranges from several hundred to 2.8 billion dollars.
“Many effective drugs were actually discovered by accident, rather than purposefully designed,” Madani said. “[Profluent’s] The ability offers humanity the opportunity to move from accidental discovery to deliberate design of our most necessary solutions in biology.”
The 20-employee Berkeley-based Profluent is backed by VC heavy hitters including Spark Capital (which led the company’s recent $35 million funding round), Insight Partners, Air Street Capital, AIX Ventures and Convergent Ventures. Google’s Chief Scientist Jeff Dean has also contributed, lending additional credibility to the platform.
Profluent’s focus in the coming months will be upgrading its AI models, in part by expanding its training datasets, Madani says, and acquiring customers and partners. He should move aggressively. Rivals, including EvolutionaryScale and Basecamp Research, are quickly training their own protein-producing models and raising huge amounts of VC cash.
“We developed our original platform and demonstrated scientific breakthroughs in gene editing,” Madani said. “Now is the time to scale up and start delivering solutions with partners that match our ambitions for the future.”
