What if AI could generate novel drugs to target any disease, overnight, ready for clinical trials?
What if it were able to design a drug that was perfect specifically for you?
But before we dive into how AI will disrupt today’s multi-trillion-dollar pharma industry, it’s worth noting how monolithic and slow today’s pharma industry actually is:
3 Gruesome Facts about Today’s Pharma Industry:
#1. Today, the global pharmaceutical market—one of the slowest, most monolithic industries to adapt—generated roughly $1.6 trillion in revenue in 2023. The top 10 pharmaceutical companies alone generated revenues of nearly $550 billion last year.
#2. At the same time, it currently costs more than $2.5 billion (sometimes up to $12 billion) and takes over 10 years to bring a new drug to market.
#3. Meanwhile, 9 out of 10 drugs entering Phase I clinical trials will never reach patients.
But all of this is about to change, unleashing what you might call “pharmaceutical abundance.”
This will happen as AI converges with massive data sets in everything from gene expression to blood biomarkers, enabling novel drug discovery that is 100-fold cheaper, 100-fold faster, and more intelligently targeted.
In today’s blog, I’ll discuss 3 companies that are using AI for better drug discovery and revolutionizing medicine: Insilico Medicine, Google DeepMind, and SandboxAQ.
Let’s dive in…
Insilico Medicine: Revolutionizing Drug Discovery with AI
Insilico Medicine is on a mission to revolutionize the traditionally slow and costly process of drug discovery by harnessing the power of AI. Founded by Dr. Alex Zhavoronkov, Insilico uses AI to rapidly identify novel drug targets, test drug candidates, and output those that are ideal for further development. (Full disclosure: My venture capital firm BOLD is an investor in Insilico Medicine.)
What sets Insilico apart is their unique approach to AI, leveraging generative adversarial networks (GANs) to accomplish with just 50 people what a typical drug company does with 5,000.
As Dr. Zhavoronkov asks:
“What if AI could find better drug targets, faster? And what if it could create new drugs that might work on these targets for us?”
Insilico’s recent milestones speak for themselves. In April 2023, the FDA cleared Insilico to begin trials of an AI-designed small-molecule treatment, which they later licensed to Exelixis. June 2023 saw the first patient dosed in a Phase 2 trial of an AI-designed treatment for idiopathic pulmonary fibrosis.
In November 2023, Insilico’s AI-powered target-discovery engine, PandaOmics, identified “dual purpose” targets for both anti-aging and brain tumor treatments.
Most recently, in April 2024, Insilico’s AI-designed drug ISM3412 received FDA IND approval, paving the way for a Phase I multicenter, open-label study in adult subjects with locally advanced and metastatic solid tumors. And in May 2024, Insilico partnered with NVIDIA to unveil a new LLM transformer, nach0, for solving biological and chemical tasks.
AlphaFold 3: Unlocking the Secrets of Life’s Building Blocks
Inside every cell are billions of molecular machines made up of proteins, DNA, and other molecules (lipids, carbohydrates). To truly understand life’s processes, we need to see how these molecules interact together.
Enter AlphaFold-3.
In May 2024, Google DeepMind unveiled this revolutionary model that predicts the structure and interactions of all life’s molecules with unprecedented accuracy. AlphaFold-3 improves prediction accuracy by at least 50% compared to existing methods, and even doubles it for some important categories.
DeepMind CEO Demis Hassabis says, “Properties of biology emerge through the interactions between different molecules in the cell. You can think about AlphaFold 3 as our first big sort of step towards that.”
Building on the groundbreaking AlphaFold-2, which has already been used by millions of researchers for discoveries in areas like malaria vaccines and cancer treatments, AlphaFold-3 goes beyond proteins to a broad spectrum of biomolecules. This leap could unlock more transformative science, from developing biorenewable materials and more resilient crops, to accelerating drug design and genomics research.
The implications for drug discovery are game-changing. As Max Jaderberg from Isomorphic Labs, a subsidiary of Alphabet founded by Hassabis, explains:
“It allows our scientists, our drug designers, to create and test hypotheses at the atomic level, and then within seconds produce highly accurate structure predictions… This is compared to the months or even years it might take to do this experimentally.”
SandboxAQ: Transforming Drug Discovery with AI & Simulation
My dear friend Jack Hidary, CEO of SandboxAQ, has highlighted the need for more effective treatments for the 7,000+ diseases that lack them.
Hidary emphasizes:
“Roughly 88% of drugs that reach clinical trials fail to make it through and gain approval.”
SandboxAQ is addressing this challenge with its unique approach combining AI and simulation. By digitally modeling biopharma compounds and simulating their interactions with molecular targets, the company generates data to develop treatments for the most challenging illnesses.
In November 2023, SandboxAQ partnered with NVIDIA to predict chemical reactions for drug discovery.
The January 2024 acquisition of Good Chemistry further strengthened SandboxAQ’s position, integrating the firm’s software into its enterprise portfolio.
SandboxAQ’s computational tools are helping to revolutionize the pharmaceutical industry by reducing the time, cost, and risk of identifying promising compounds.
As Nadia Harhen, General Manager of Simulation and Optimization at SandboxAQ, explains, “We see hundreds and hundreds of clinical trial failures, and they’re labeled undruggable because no one’s been able to drug them yet because it’s complicated—it’s too complex to drug…That’s what we build specialized tooling to tackle, and that specialized tooling we’re going to make available on this platform.”
During my upcoming Longevity Platinum Trip, I’ll be featuring in-depth conversations with several leading companies and scientists in AI-powered drug discovery, including:
- Rubedo Life Sciences – Co-Founder & CEO, Marco Quarta, PhD: A biotech company that leverages machine learning and single-cell RNA sequencing to develop innovative therapeutics for age-related diseases such as psoriasis and scleroderma.
- Verge Genomics – Founder & CEO, Alice Zhang: An AI-driven biotech company that is revolutionizing drug discovery with their platform, CONVERGE®, particularly for debilitating conditions like amyotrophic lateral sclerosis (ALS).
- BioAge Labs – Co-Founder & CEO, Kristen Fortney, PhD: A pioneering biotech company that uses AI to enhance drug discovery and development, with a primary focus on therapies for aging and metabolic diseases. BioAge Labs is currently focusing on metabolic disease: an oral drug aimed at enhancing weight loss and improving metabolic health (Azelaprag).
- Phenome Health – Lee Hood, MD, PhD (CEO) & James Yurkovich, PhD (Chief Innovation Officer): A cutting-edge biotech company that combines multi-omics, AI, and large-scale data to combat aging. The company is led by pioneers in the fields of systems biology research and personalized medicine and has established a partnership with the Buck Institute for Research on Aging.
Why This Matters
Inefficient, slow-to-innovate, and risk-averse industries (like pharma) will be disrupted by leaders and tech coming from outside their normal ranks, in this case from today’s AI geeks.
If you are like me and desire to maximize your healthspan (and longevity), then you’re likely to realize that optimizing your sleep, diet, and exercise can only get you so far… perhaps to a healthy 90 or 100 years old, but it’s unlikely to get you further.
To get beyond 100 and strive towards “Longevity Escape Velocity,” we’re going to need to understand (on a molecular level) why each of us ages and design the drugs that slow that process down and perhaps even reverse it.
For me, AI is the most important technology speeding us toward an extended healthspan future, and companies like Insilico, DeepMind, and Sandbox AQ are leading that charge.
