Photo Credit: M. Richter from Pixabay
The Small Molecule Drug Discovery Market is estimated to grow at 9% CAGR from $74.08 billion in 2024 to reach $175.41 billion by 2034. San Francisco-based Atomwise is one player that is making use of its AI capabilities to make a mark in the industry.

Atomwise’s Offerings
Atomwise was founded in 2012 with the intention of finding cures for common and orphan diseases that are too expensive and time-intensive for pharma companies to find a cure for. It was one of the first companies to deploy the power of AI’s convolutional neural networks and apply that to drug discovery for molecular recognition with the objective of finding the right drug.For a drug to work, it needs to stick to the disease protein to slow down its growth. The drug needs to ensure that it doesn’t harm any of the other necessary proteins. Atomwise’s AI platform, AtomNet, uses a deep convolutional neural network that excels at understanding complex concepts as a combination of smaller and smaller pieces of information. Convolutional networks have helped improve AI capabilities for seemingly routine activities like image classification and speech recognition.Within drug discovery, AtomNet uses that capability to recognize essential chemical groups like hydrogen bonding, aromaticity, and single-bonded carbons by studying vast quantities of target and ligand data. Using its learning, AtomNet is also able to predict the results of physical experiments. It not only identifies promising compounds but can also make predictions for virtually synthesized novel molecules that have never physically existed. By incorporating protein structure in its analysis, it has been able to provide better accuracy and refine predictions.According to Atomwise, AtomNet identifies compounds at a rate 10,000 times higher than that of physical screens, or 100 times faster than ultra-high-throughput screening, enabling it to screen more than 10 million compounds each day. When the platform was being built, it went through a benchmark test called DUD-E, which was designed by University of California, San Francisco. The benchmark uses more than 1 million predictions made by the systems and compares them to historical results. When tested, AtomNet achieved the best results of any structure-based algorithm.AtomNet is being used in projects involving hit discovery, lead optimization, off-target toxicity, selectivity, and cross-species activity. Besides an increase in discovery productivity, Atomwise’s solution also provides greater understanding of the toxicity, side effects, mechanism of action, and efficacy of a drug, at a much earlier stage in a drug pipeline, thus reducing cost and risks for drug developers.Atomwise has had a successful run so far. It has several partnerships with pharmaceutical companies and academic groups. Within three years of its launch, it was working with IBM and other institutions on over a dozen projects tackling diseases like multiple sclerosis and ebola. It had also entered into collaborations with Merck, Notable Labs, Hansoh Pharma, Eli Lilly, and Harvard Medical School, to name a few.In more recent years, Atomwise has been migrating from a more collaborative drug discovery model to one where it focuses on building a pipeline of drug assets internally. Last year, it announced that its small molecule focused on TYK2 inhibition had been nominated as its first AI-driven development candidate. TYK2 is a key mediator in cytokine signaling pathways that could address a broad range of immune-mediated inflammatory conditions.

Atomwise’s Financials
Atomwise earns revenues through its partnership agreements on drug discovery, along with licensing for Atomnet and consulting for its services. It is privately held, and does not disclose its financial performance. Reports suggest that it has raised $173 million in funding so far in seven rounds of funding at an undisclosed valuation. Its investors include Monsanto Growth Ventures, DCVC (Data Collective), B Capital Group, Y Combinator, Khosla Ventures, DFJ, Baidu Ventures, Tencent, and Dolby Family Ventures.There are several players within the AI-based drug discovery market. A win from any of these players will help seal the importance of AI in the industry.More By This Author:AI Drug Discovery Startups: Could Insilico Medicine Release The First Drug To Be Discovered By AI?
AI Drug Discovery Stocks: Recursion Pharmaceuticals Has A Long Road Ahead
Exa Wants To Be The Google For AI