SandboxAQ Launches Biopharma Division for Faster Treatment Development

SandboxAQ has launched a biopharma molecular simulation division, AQBioSim. The company combines AI and quantum technology, and the new spinout aims to use this to decrease the time and cost of developing new therapies and de-risk drug portfolios before entering preclinical and clinical stages. 

New methods of simulation could help biopharma companies achieve breakthroughs in treatments for conditions including cancer, Alzheimer’s and Parkinson’s. The organizations AQBioSim is working with include  AstraZeneca, Sanofi and UC San Francisco’s Institute for Neurodegenerative Diseases. 

In this Q&A, the general manager for SandboxAQ’s simulation and optimization group Nadia Harhen explains how the technology works and the significance of the move.

Enter Quantum: Why did SandboxAQ create AQBioSim?

Nadia Harhen: It’s the culmination of talking to many top industry executives and research leaders and computational chemists from large and small organizations all over the biopharma industry.

The number of drugs being approved today is fewer than in the last 10 to 14 years. We have a thesis around this, which is that the drugs that are being approved today are harder to approve because companies have taken all of the low-hanging fruit. They have exhausted all of the easy lock a target into a pocket kind of drugs.

All that's left are these difficult-to-treat targets the undruggable diseases, and these are the types of diseases that plague us at the end of our lives – the cancers, the neurodegenerative diseases. These are tough because it's not enough to just buy into the protein, you have to do something else to it.

The other one is that all the tools and methods that they’ve used in the past no longer work. That's where we come in. We have taken a fresh look at this problem; no one else is applying quantum AI quite in the way that we are talking about.

There’s not a lot of data about these undruggable diseases, so AI is not going to help you. So if you start with a quantum first approach, where you're looking at the quantum mechanical interactions and molecules using quantum equations given to us 100 years ago, by Einstein and Schrödinger and others, calculating the physics yields very accurate results that you can then use to train an AI model on.

That's our unique combination of doing things, and then once you do that, you get a very accurate representation of how these molecules are interacting in the presence of blood or heat or just other physics attributes that can be calculated in silico on classical hardware.

What stages of the drug and treatment discovery process can AI quantum (AQ) technology help with?

It can be applied at all stages of the drug discovery through to the clinical trial process. We work to connect drug discovery to development; that's the primary area in which we impact the other part of what we focus on.

Not only do we impact the virtual screening has to lead optimization in the early stages of drug discovery, but then in silico, hypothesis testing using AI methods. That one has potential to span across the different phases of drug discovery, even in clinical trials where you're trying to figure out why did something fail? Which population does it work better on and what is the mechanism of reaction that is leading to some of this you know, adverse events that we're seeing?

How do you work with your clinical clients?

We work in a very collaborative manner. Our clients are the experts at what they do; they have years and years of not only drug expertise in a particular disease area but also the technical know-how they have worked with a molecule.

We help them figure out and answer hard questions, like ‘What is the mechanism of action’ or ‘We’re struggling with getting a particular adverse reaction out of this drug’. We work on very specific, hard-to-treat problems so that we can deliver them back either an optimized molecule or help answer some of those tough questions.

How will AQ methods eventually help clinical outcomes for patients?

We want to put out more accurate drugs on the market that are safer, with minimal adverse reactions, and effective. We also want better efficacy by designing an optimal molecule to get the intended effect.

If we can do this faster, we believe that we can generate more molecules and put them out into the ecosystem, and eventually deliver faster, safer drugs to the end patient. Hopefully, because we have all of these drugs, we can pass on those cost savings to the customer. We want to do good and leave a lasting legacy, outside of the company's existence.

This will include undruggable diseases – Alzheimer's falls into that category, MS, Parkinson's – a lot of neurodegenerative diseases, autoimmune diseases and cancers fall into that category. Many of these folks get end of life are really hard to treat.

What’s next for SandboxAQ and AQBioSim?

We are not only starting in drug discovery, but we are also looking at material science discovery and impacting financial services with simulation tools. We believe that in the next few years, all businesses will need to use SandboxAQ simulation tools to develop new molecules.

We’ve seen how the folks that we're collaborating with are leapfrogging their competitors in being able to take years off of their clinical trial time. If you’re a big pharmaceutical company developing a blockbuster drug, every year you take off represents a billion dollars in revenue.

The value is really clear and we're hoping that we can pass that on to other pharmaceutical companies and smaller institutions that are developing drugs that don't have the resources and tools. We want to give them really good tools intended to develop better molecules and then pass that technology on to materials and other and other sectors.

This article first appeared in IoT World Today's sister publication Enter Quantum.