In between my stint in particle physics and career in AI, I enjoyed 12 years in drug design at CSIRO (Australia’s Commonwealth Scientific & Industrial Research Organization) and I remember a fascinating piece of work which came out at that time and got talked around the lab coffee room for a few weeks.

It was a simple idea but perfectly suited to the 3D structure-determining technique we were using in our lab: NMR (Nuclear Magnetic Resonance), a sister technique to X-ray Crystallography, and related to MRI, except works on test tube samples and at a near-atomic level rather than an anatomical level.

The idea that I think originates with Sergio Rotstein & Mark Murcko’s paper in 1993 was that with NMR you can detect weak interactions, and their nature, of small molecules with target proteins important in disease. Then the idea was tha tyou could tether multiple such weak interactors together to make a highly potent drug.

Well. in the last ten years this is really gaining traction. Related methods are discovering new drugs (eg JAK kinase inhibitors) and maybe even an entire new method of drug discovery.

And now startups (eg Vividion Therapeutics), incumbents (eg ZoBio) and cooperative research centres are popping up all over the place.

Challenge

The challenge for drug discovery is to find small molecules that are active, highly potent, orally active and non-toxic and wihtout side-effects.

That’s a tall order.

And some of the most important cancer targets like Myc and Raf have flat interfaces and little for a ‘lock and key’ molecule to lock on to.

Some call them ‘undruggable’.

But through methods like fragment-based drug design (FBDD) and tethering smaller low-interaction molecules together we might get there. Two small weak-interactions multiplicatively amplify. That’s how it works.

And using Ai and structural bioinformatics could filter in the best candidates for testing.

The market could be huge for such discoveries. Think a big chunk of the $160 billion cancer pharmaceutical market. Or more.

DISCLOSURE
I am founder of a related drug discovery company: CancerONE P/L.

Image credit: Paul Pallaghy | CancerONE P/L