I: Insilico Medicine 10 Years of AI Drug Discovery Innovation
I: Insilico Medicine 10 Years of AI Drug Discovery Innovation
Revolutionizing Drug Development
“This first drug candidate that’s going to Phase 2 is a true highlight of our end-to-end approach to bridge biology and chemistry with deep learning.
This is a significant milestone not only for us, but for everyone in the field of AI-accelerated drug discovery.”
By Alex Zhavoronkov, PhD, Founder and CEO of Insilico Medicine
Latest News On Insilico Medicine
🐞 Using generative AI, Insilico Medicine discovers new class of Polθ Inhibitors for BRCA-deficient cancers.
🐞 Insilico Medicine to sponsor ARDD 2024, the world's largest conference on aging research in the biopharmaceutical industry.
🐞 Insilico Medicine is collaborating with SRW Laboratories on longevity research with substantial laboratory robotics validation.
Insilico Medicine 🎩
Insilico Medicine founded by Alex (Aleksandrs Zavoronkovs) Zhavoronkov, is a company that doesn’t need introductions. The legendary Insilico Medicine—together with Exscientia, Atomwise, Recursion Pharmaceuticals, Iktos and many more—they are all considered global leaders in the AI drug discovery space, something like Neil Armstrong and Yuri Gagarin the two greats from spaceflight and space exploration. Back to earth now, these companies have been exploring the chemical space for “drug discovery missions” doing actually great work!
Insilico Medicine (Science Park, Hong Kong, New York) was effectively born 🐣 in 2014 at NVIDIA GTC—a global AI conference with a focus on DL for drug discovery—and combines genomics, big data analysis and DL for in silico drug discovery. Insilico’s AI platform, Pharma.AI, it’s a fully integrated drug discovery software suite that can:
discover new targets with PandaOmics, enabling multi-omics target discovery and deep biology analysis to considerably reduce required time,
PandaOmics now uses 1.9 trillion data points drawn from over 10 million samples (such as microarrays, RNA sequencing, and proteomes) and over 40 million documents (including patents, grants, publications, and clinical trial reports).
design new drugs with Chemistry42, to find novel lead-like molecules through automated ML de-novo drug design and scalable engineering platform,
predict the outcomes of clinical trials with inClinico, enabling to predict clinical trials success rate and recognize the weak points in trial design, while adopting the best practices in the industry, and
in particular for biologics they offer Generative Biologics, an advanced AI-powered platform that is capable of designing and optimizing different types of biologics including peptides, nanobodies and antibodies tailored for specific targets.
In a new paper in Nature Biotechnology (A small-molecule TNIK inhibitor targets fibrosis in preclinical and clinical models) Insilico Medicine describes step-by-step the process that undertook to develop INS018_055—a small-molecule TNIK inhibitor that is currently in phase 2 trials for the treatment of the lung disease idiopathic pulmonary fibrosis—by using its proprietary AI platform, PandaOmics, to both identify a target and come up with a drug candidate to treat the disease.
INS018_055 (Idiopathic pulmonary fibrosis), a protein kinase inhibitor like ISM042-2-048, is the first drug discovered and designed using AI to reach the phase I clinical trial milestone.
Moreover, Insilico just presented their first demo of the PreciousGPT for aging research—Insilico's lineup of AI models aimed at enabling digital omics experiments—that includes:
Precious1GPT: a transformer-based model with aging clock functionality for aging-related pathology research with a simple transformer regressor and transformer classifier architecture,
Precious2GPT: a compound model combining a transformer and a diffusion architectures with omic data generation capabilities, and
Precious3GPT: a genuinely multimodal transformer-based model trained to emulate the workflow of case-control studies, with an emphasis on chemical perturbations. Precious3GPT is a
multimodal,
multi-omics (proteomes, RNA sequencing and DNA methylation data),
multi-species,
multi-tissue, and
multi-tasking
transformer for drug discovery and aging research ("biology in time"). For more: Precious3GPT.
Insilico’s AI pharma collaborations
Insilico’s AI pharma collaborations so far have been the following:
🏏 In 2017, Insilico Medicine established a collaboration with GSK (LON: GSK) to discover novel biological targets and molecules.
In 2020, Insilico launched a preclinical research program focused on finding new treatments for brain cancer, and has brought on the former global program head of GSK’s computer-aided drug discovery unit to help run it.
🏏 In November 2020, Insilico Medicine announced that Merck KGaA (ETR: MRK), Darmstadt, Germany will be the first launch partner for its flagship generative chemistry AI platform Chemistry42, and that Merck will integrate Chemistry42 into its discovery pipeline to facilitate rapid and effective drug design. Chemistry42, core part of Insilico's Pharma AI drug discovery process, is a user-friendly platform that bridges AI and ML methods with domain expertise in the fields of medicinal and computational chemistry, for the design of novel small molecules with desirable physicochemical properties.
🏏 In 2020, it was announced that Insilico Medicine teamed up with Pfizer (NYSE: PFE) in a deal that will see Pfizer leverage Insilico's technology to identify drug targets for a variety of diseases. Pfizer by using Insilico's technology can speed up drug discovery—and possibly slim down drug development costs in the process.
🏏 On January 11, 2022, Shanghai Fosun Pharma—focusing on major therapeutic areas, including oncology, immunology, the "4 hypers": hypertension, hyperlipidemia, hyperglycemia and hyperuricemia and their complications, as well as central nervous system—and Insilico Medicine entered into a collaboration agreement to advance the discovery and development of four biological targets, as well as the co-development of Insilico's QPCTL program (glutaminyl-peptide cyclotransferase-like protein). Under this agreement, Insilico will receive a total upfront payment of $13M.
Interestingly, in under 40 days since the collaboration announcement, the two companies nominated their first preclinical drug candidate ISM004-1057D, a potential first-in-class small molecule inhibitor that targets QPCTL.
🏏 On March, 24, 2022, Insilico Medicine announced it has entered into a strategic collaboration with EQRx, a company committed to developing and delivering innovative medicines to patients at radically lower prices. Under this agreement, Insilico will apply its Pharma.AI platform to advance de novo small molecule design and generation with EQRx's clinical development and commercialisation expertise.
In November 9, 2023, Revolution Medicines—a clinical-stage oncology company developing targeted therapies for RAS-addicted cancers—announced it had closed its acquisition of EQRx. Through the acquisition, Revolution Medicines expects to add approximately $1.1 billion in net cash proceeds to its balance sheet, after estimated post-closing EQRx wind-down and transition costs.
🏏 Moreover, in 2022 Sanofi (EPA: SAN) joined forces in a collaboration with Insilico Medicine (signing a strategic research worth up to $1.2 Billion). Under the terms of the agreement, the collaboration will leverage Insilico Medicine’s AI platform, Pharma.AI, to advance drug development candidates for up to six new targets.
🏏 In September 12, 2023, Exelixis—an oncology company innovating next-generation medicines—and Insilico Medicine announced that they have entered into an exclusive license agreement granting Exelixis global rights to develop and commercialize ISM3091, a potentially best-in-class small molecule inhibitor of USP1, which has emerged as a synthetic lethal target in the context of BRCA-mutated tumors.
Accordingly, Insilico granted Exelixis an exclusive worldwide license to develop and commercialize ISM3091 (and other USP1-targeting compounds) in exchange for an upfront payment to Insilico of $80M. Exelixis located in Alameda, California, is the producer of Cometriq, a treatment approved by the FDA for medullary thyroid cancer with clinical activity in several other types of metastatic cancer.
🏏 In 2023, Insilico and Novo Nordisk (CPH: NOVO-B) signed a collaboration agreement that will help the Danish pharma identify targets against liver fibrosis. Novo Nordisk will utilize Insilico’s AI target discovery engine, PandaOmics, in conjunction with its own drug research and development expertise, to yield new target hypotheses of interest to Novo Nordisk.
🏏 In January 2024, Menarini Group and Insilico Medicine (and Stemline Therapeutics, Inc. a wholly-owned subsidiary of the Menarini Group) entered a global exclusive license agreement for novel KAT6 Inhibitor for potential breast cancer treatment and other oncology indications. The novel molecule was designed by Insilico’s R&D team with the help of its Pharma Generative AI platform, to inhibit KAT6A and block endocrine receptor (ER) at the transcriptional level, giving it the potential to overcome resistance to endocrine therapies due to mutation or ligand-independent constitutive activation of ER.
Under the terms of the agreement, Stemline will provide a $12M upfront payment to Insilico. The combined value of the deal, including all development, regulatory, and commercial milestones, is over $500M, followed by royalties up to double digits.
💎 On June 29, 2023, Insilico Medicine Filed For Hong Kong IPO.
“Insilico Medicine did not disclose details of its IPO in its filing to the Hong Kong stock exchange on Tuesday, although a report by local newspaper South China Morning Post said it’s planning to raise $200M, citing unnamed sources. The company had previously filed confidentially for a U.S. IPO to raise around $300M, Bloomberg News reported in November 2021. Insilico Medicine didn’t immediately respond to a comment request.”
Insilico Medicine: Time and Money
2023 was undoubtedly an important year for AI Drug Discovery and Insilico Medicine (AI’s evolving role in drug discovery and development in 2023) since:
In June 2023: Insilico Medicine’s AI-drug entered phase 2 study.
INS018_055 was the first AI-discovered drug to reach phase 2 trials and with traditional methods the drug candidate would have cost more than $400M and would have taken up to six years to develop. But with generative AI, it reached those objectives at one-tenth (1/10) of the cost and in one-third (1/3) of the time!
In August 2023: Generative AI tool boasts 79% accuracy in predicting clinical trial outcomes.
Insilico Medicine announced that its generative AI tool InClinico (that uses transformer-based AI models and multimodal data sources, including text, omics, trial design and drug properties) reached a significant breakthrough in predicting clinical trial outcomes, after the company trained the system on more than 55,600 phase 2 clinical trials.
Insilico also noted that the platform could serve as a valuable tool for investors, supporting a 35% 9-month return on investment in a virtual trading portfolio.
However, in the article “AI in Drug Discovery in China, Hype or Hope?” published by EqualOcean, the author points out to the importance of valuable clean data (literature, large public databases and proprietary databases) but also to the importance of the experimental validation of AI algorithms in this data-intensive industry. He is actually saying that apart the problem we have with the dirty data (such as skewed data sets and biased data that the scientists need to explain that to the algorithms), there is also the problem with the experimental validation of the AI algorithms, that is impeded by two aspects:
firstly, there is no universal protocol to follow about what correct validation should be, in another way to say, there is no standard in parallel when AI delivers new compounds, and
secondly, the problem lies in the progress of validation itself.
In other words, we use data to train an algorithm and generate novel AI compounds, and for that reason the data set we use is intrinsically related to the numerical performance and the prospective performance of the method/model itself. Therefore, the so-called validation of the AI algorithms could just be replication of the construction process. Thereby, the scientists need to lead their own way to validate their AI algorithms/AI-generated molecules, leveraging their expertise. At the end of the day, wet-lab experiments, biological assays and clinical trials are also needed to ensure that the AI generated molecules are active both in vivo and in vitro.
And that requires time as Alex Zhavoronkov PhD, Founder and CEO of Insilico Medicine told EqualOcean:
“Unlike in other areas where AI generates pictures, music or text, you get validation almost immediately because you get almost immediate feedback by looking at the picture, listening to music and reading the text.
In biology, it is not like that and you have to wait”.
In any case apart the problems with dirty data and experimental validation of algorithms, in China ⛩️ for example effort has been made in AI drug discovery and there have been some demonstrated successful cases showing the validation of AI algorithms in drug discovery, like with the following companies: Accutar Biotech (AC0682, AC0176), Galixir (unknown), MindRank.AI (MDR001, EMDR001), Nutshell Therapeutics (NST001, NST004) and off course Insilico Medicine (ISM001-055, ISM012-077, ISM004-1057D, undisclosed).
NVIDIA Corporation (NVDA) and Insilico Medicine
🐞 Insilico was a premier member of NVIDIA Inception, a free program that provides cutting-edge startups with technical training, go-to-market support and AI platform guidance.
🐞 The company uses NVIDIA Tensor Core GPUs in its generative AI drug design engine, Chemistry42, to generate novel molecular structures—and was one of the first adopters of an early precursor to NVIDIA DGX systems in 2015.
🐞 Moreover, at Insilico they’re now using NVIDIA BioNeMo to accelerate the early drug discovery process with generative AI.
From Quicker Cures: How Insilico Medicine Uses Generative AI to Accelerate Drug Discovery
by Renee Yao (that leads global healthcare AI startups at NVIDIA)
Insilico Medicine and Quantum Computing
Insilico Medicine, announced in a study just published that it successfully combined generative AI and quantum computing to accelerate drug discovery in order to explore lead candidate discovery in drug development. In particular, they demonstrated the potential advantages of quantum generative adversarial networks in generative chemistry, by an implicit GAN for small molecular graphs, with a variational quantum circuit (VQC) as the noise generator.
Building on these findings, Insilico scientists plan to integrate the hybrid quantum GAN model into Chemistry42, the company's proprietary small molecule generation engine, to further accelerate and improve its AI-driven drug discovery and development process.
Insilico’s Valuation 🏌
⛳️ Insilico Medicine was valued at approximately $895M after raising $95M in July 2022, from the leading Chinese healthcare-focused firm Qiming Venture Partners and Singapore-based billionaire Eduardo Saverin's B Capital.
⛳️ On August 10, 2022, Insilico Medicine raised additional $35M from Prosperity7 Ventures, a global VC fund by Saudi Arabia’s giant Aramco.
⛳️ Insilico Medicine has raised a total of $401.3M.
Robots
The company also has plans to launch a fully automated AI-driven robotics lab for drug discovery.
In particular, Insilico Medicine launched a 6th generation Intelligent Robotics Drug Discovery Laboratory, in Suzhou BioBAY Industrial Park in December 2022, which is a fully automated AI-powered robotics laboratory that performs target discovery, compound screening, precision medicine development and translational research.
So far, the so-called “5th generation robotics laboratories” were full automation labs with no human bias or influence, connecting multiple processes and generating high-quality data that can be used for ML. Insilico's lab (the 6th generation) takes this process one step further by incorporating AI in decision-making and by integrating AI with automation, robotics and biological capabilities to enable a new generation of intelligent robotic labs.
The Intelligent Robotics Lab forms a closed loop by combining Pharma.AI with fully automated biological experiment functional modules. In particular, PandaOmics of Pharma.AI predicts novel targets for specific diseases, the robotics lab will conduct early-stage drug discovery experiments (like target validation, high-throughput screening, hit-to-lead optimization, lead-to-preclinical candidate confirmation and translational research) through six fully automated modules including sample management and quality control, compound management, automated cell culture, high-throughput screening, high-content cell imaging and next-generation sequencing. All data generated by the Intelligent Robotics Lab will complement and expand Insilico's existing data resources.
“Insilico's robotics lab has an AI brain, an automated machine body, and the limbs of various complex robots.
The AI brain has been trained and verified through Insilico projects and has learned from years of experience collaborating with global pharmaceutical companies.
It can carry out systematic learning based on the information provided, and assist in the decision-making of target discovery and identification.
In the Intelligent Robotics Lab, the AI brain will propose potential targets and design automated experiments and workflows based on experimental results."
Alex Zhavoronkov, PhD, Founder and CEO of Insilico Medicine
“We want companies and fellow cancer researchers to see the remarkable value of AI in target and biomarker discovery and de novo drug design in developing new cancer drugs, driven by a team of cross-trained AI experts and drug developers. We hope they will leave with assurance that Insilico’s AI platform can be used to create highly optimized, potent, and efficacious molecules that can serve as potentially best-in-class therapeutic options in treatment-resistant cancers and as promising candidates for partnering.”
Michelle Chen, PhD, chief business officer at Insilico Medicine, told GEN
Insilico Medicine and Innovation
Moreover, regarding the study published in Nature Biotechnology that presents the entire journey of INS018_055 from AI algorithms to Phase II clinical trials for the first time, Insilico developed a PaperGPT system based on ChatGPT-4 Turbo and internal LLM that provides answers related to the paper via chat functionality. For example:
Q: “How notable is this study and what is important about it?”
A: “Thank you for your question regarding the significance and importance of our study. The importance of this study lies in several key aspects:
Innovative Methodology: We introduce a state-of-the-art drug-discovery methodology that leverages AI, specifically designed to enhance the efficiency and efficacy of identifying potential therapeutic compounds.
Clinical Relevance: The paper presents the results of three meticulously designed clinical trials that adhere to the highest ethical standards of good clinical practice, according to the International Council for Harmonisation (ICH) Harmonized Guideline E6(R2). These trials provide crucial data on the safety and potential efficacy of the AI-predicted compound INS018_055, marking a significant step forward in the translation of AI-predicted findings into tangible clinical benefits.
Transparency and Reproducibility: We have made a conscious effort to ensure the transparency and accessibility of our research data and methodologies. All associated data, including full clinical study protocols for all three trials, are provided in the Supplementary Information.
Addressing the Need for Novelty: In light of the documented challenges in drug development, including high attrition rates and the increasing cost of in vivo studies, our study represents a novel approach that could potentially streamline the drug discovery process and reduce associated costs. By demonstrating the safety of the AI-predicted compound in phase I studies and providing a comprehensive framework for AI-enabled drug discovery, we offer new avenues for exploring therapeutic opportunities within the human genome.”
To conclude, Insilico Medicine was named Top Biotech Company in Fast Company’s 2024 World’s Most Innovative Companies List and has just appointed Keith Mikule, PhD, as Vice President (VP) of Business Development. Based in Boston, Dr. Mikule will work closely with Alex Zhavoronkov, PhD, founder and CEO, Feng Ren, PhD, co-CEO and Chief Scientific Officer, and Michelle Chen, PhD, Chief Business Officer, to drive business development activities and operation strategies to support Insilico's continued growth.
For the second and third part of this newsletter: