Unnatural Products Aims to Unlock Drug Discovery With Machine Learning

By Brent MacDonald – Principal @ Rising Tide VC

Unnatural Products mission is solving the problem of ‘undruggable’ proteins—the roughly 80% to 90% of proteins in the body that have eluded drug makers. Some proteins are inaccessible because they lack a binding “pocket” where drugs can dock. What is more, traditional drug development approaches are rapidly becoming exhausted. This leaves a large range of drug targets left ‘undruggable’ and are beyond the reach of biologics and too complex for small molecules.

In many ways our understanding of human biology, particularly in the role of certain drivers of cancer, has outstripped our chemistry capabilities” – Cameron Pye, Co-founder & CEO Unnatural Products

Unnatural Products (UNP) is a drug discovery startup pairing AI with chemistry to create treatments for drug targets that are not amenable to current therapeutic strategies. The Company has developed a Machine Learning platform to engineer passive cell permeability into larger molecules, capable of binding these undruggable targets.[1] UNP founders are scientists from the University of California and are focused on one of nature's solutions to this problem, macrocycles, to expand the chemical toolbox. They have also worked with both pharma and academia to develop macrocycle programs. Macrocycles, which are found in nature, bridge the gap between small molecules and biologics by transcending the traditional boundaries of cell permeability to access deep, complex targets with high specificity. They are molecular structures that contain one or more rings of at least 12 atoms with the unique characteristic that they combine the benefits of large biomolecules, such as high potency and exquisite selectivity, with those of small molecules, including reasonable manufacturing costs, favorable pharmacokinetic properties, including oral bioavailability, ease of administration and lack of immunogenicity.

Today, there are a number of companies focused on discovering macrocycles for specific drug targets (Hits). However, the majority of these companies face the same challenge once they have identified a hit: cell permeability. UNP’s machine learning platform optimizes macrocycle design and predict their passive permeability into cells. The Company is currently focused on oncology, though their technology can be applied to a wide variety of urgent clinical needs such as inflammatory diseases.

Why We Are Excited:

Macrocycles have been an active area of development for the past 10+ years, but many compounds have stalled due to poor permeability. UNP’s platform to optimize macrocycles structures to improve passive permeability without sacrificing target-binding affinity has the potential to enable the development of drugs for many ‘undruggable’ targets. Additionally, once compounds have been optimized to permeate cells, UNP has the ability to build a unique dataset around compound performance that many others have not yet had access to. UNP’s unique platform closely marries prediction with experiment. This rapid, intelligent learning loop has allowed UNP to solve key challenges that have previously been the roadblock in macrocycle development.

A lot of the high hanging fruit in drug discovery has been called ‘undruggable,’ Unnatural Products is working to change that.

[1] Passive permeability is a type of membrane transport that does not require energy to move substances across cell membranes. The rate of passive transport depends on the permeability of the cell membrane, which, in turn, depends on the organization and characteristics of the membrane lipids and proteins.