Deep Apple

About Deep Apple

Deep Apple Therapeutics is a biotechnology firm focused on accelerating the discovery of novel small molecule drugs. The company was created and incubated by the life sciences venture capital firm Apple Tree Partners (ATP), officially launching in December 2023 with a $52 million Series A commitment from ATP. The company was founded by a team of academic leaders whose expertise forms the core of its discovery engine. This includes Georgios Skiniotis, Ph.D., of Stanford University, an expert in cryo-electron microscopy (cryo-EM) and G protein-coupled receptor (GPCR) structural biology; Brian Shoichet, Ph.D., a pioneer of virtual screening from the University of California San Francisco (UCSF); and John Irwin, Ph.D., also from UCSF, known for creating the ZINC virtual library of over 10 billion synthesizable compounds.

The company's business model centers on leveraging its proprietary platform to discover and develop drug candidates, which it then partners with larger pharmaceutical companies to commercialize. This strategy is exemplified by a June 2025 research collaboration and license agreement with Novo Nordisk to develop oral small molecule therapeutics for cardiometabolic diseases, including obesity. Under this agreement, Deep Apple is responsible for the discovery and optimization phases, with Novo Nordisk handling development, manufacturing, and commercialization. The deal includes an upfront payment and makes Deep Apple eligible for up to $812 million in research and milestone payments, plus potential royalties.

Deep Apple's platform integrates three key technologies to shorten the drug discovery timeline from years to less than 12 months. First, it uses ensemble cryo-EM to study the various conformations and transient binding pockets of protein targets, particularly integral membrane proteins like GPCRs. Second, this structural data feeds into deep learning models that analyze the raw cryo-EM data to build higher-quality 3D maps and identify fleeting pockets that could be drugged. Finally, the company performs virtual screening, or docking, of multi-billion compound libraries against these identified targets using its proprietary algorithm, Orchard.ai™. This *in silico* process allows the company to test billions of compounds in hours and then synthesize only the most promising candidates for wet-lab validation, a feedback loop that continuously improves the predictive power of its models. The company is initially focused on GPCR modulators for metabolic disorders, inflammation, and immunology, and has already initiated programs in obesity and inflammatory disorders.