Big investments in artificial intelligence (AI) don’t guarantee big wins. This point was emphasized by Niven R. Narain, the president and CEO of BPGbio president and CEO Niven R. Narain, PhD, in an editorial he wrote for Forbes.1 After noting that there has been a cumulative investment of $60 billion in AI-driven drug development,2 Narain offered a sobering analogy. He suggested that a drug developer employing AI might fail to introduce blockbuster drugs, just as a soccer club fielding superstar Lionel Messi might fail to score game-winning goals.

“The problem isn’t AI itself—it’s the approach most companies have taken to its use,” Narain continued. “The underlying issue for pharma companies using AI lies in the quality of the inputs and how AI models are used. Think of AI as the teamwork and coaching that enable players to reach their full potential.”

To stretch the analogy further, there’s no shortage these days of AI-based “teams” eager to dominate the field. But as in any other segment of drug development, these teams—or companies—will have to develop champions and win championships—by advancing their candidates through preclinical studies and clinical trials, generating data that is solid enough to gain regulatory approvals for therapies that reach the market and ultimately treat patients.

To help our readers monitor companies of this sort, GEN has prepared this A-List. It is, in fact, GEN’s first A-List devoted to AI-based drug discovery and development companies. It includes 10 companies that are broadly acknowledged as leaders in the AI field. The aforementioned BPGbio is the third company in the list. (Six additional companies are listed in the online version of this article. They appear in a separate section for companies most notable for pipeline or platform progress.)

This A-List does not encompass AI companies that build platforms and technologies rather fill their own drug development pipelines, however substantial these companies may be. Such companies include Nvidia, the Silicon Valley microprocessing giant. Nvidia has grown its market capitalization to $3 trillion and beyond by catering to multiple industries, including the life sciences industry, where the company recently introduced reference workflows for drug discovery companies.

Another company not encompassed by this A-List is Palantir Technologies. Its Foundry platform combines data processing, analytics, and machine learning to power fundamental and translational research. Customers include healthcare providers, organizations, and systems. The company also works with the NIH in the United States, and with the NHS in the United Kingdom.

 

Anima Biotech

After forming in 2018, Anima Biotech began lining up big-name partners such as Eli Lilly and Company and Takeda Pharmaceutical. Anima’s newest partner is AbbVie, which last year agreed to collaborate on the discovery and development of mRNA modulators for three targets across oncology and immunology. In this partnership, AbbVie is using Anima’s mRNA Lightning.AI platform, which images hundreds of cellular pathways in both healthy and diseased cells to train disease-specific AI models, and which uses neural networks to help these models distinguish between healthy and diseased cells and identify dysregulated pathways. These pathways are then analyzed to uncover novel targets backed by experimental validation. Anima has disclosed a pipeline of 20 programs, all preclinical, in neuroscience as well as oncology and immunology.

Atomwise

AI was hardly a household phrase in 2020 when Atomwise inked collaborations with biopharma giants Bayer, Bridge Biotherapeutics, Hansoh Pharma, and GC Pharma. Two years later, Sanofi paid Atomwise $20 million upfront to launch an up-to-$1 billion-plus collaboration designed to use its AtomNet platform to pursue up to five drug targets. In April, Atomwise published results from a 318-target study highlighting AtomNet as a viable alternative to high-throughput screening. AtomNet identified structurally novel hits for 235 of 318 targets evaluated in the study, which involved collaborations with over 250 academic laboratories across 30 countries.

BPGbio

At BPGbio, much activity is focused on BPM31510, a drug-lipid conjugate nanodispersion that targets cancer energy generation. Indeed, 6 of BPGbio’s 10 disclosed pipeline programs focus on developing formulations of BPM31510. One of these programs, BPM31510IV, has received orphan drug designation for glioblastoma multiforme and pancreatic cancer. Another program, BPM31510T, has received orphan drug designation for epidermolysis bullosa, as well as rare pediatric disease designation for primary CoQ10 deficiency.

At the recent MitoCon conference, BPGbio presented data from a collaboration with Columbia University’s Quinzii laboratory. Using BPGbio’s Quinomics technology, the collaborators generated data showing that BPM31510T outperforms traditional oral CoQ10 in treating mitochondrial dysfunction.

Earlier, BPGbio and the University of Oxford agreed to partner on advancing novel protein degradation technologies, starting with validating BPGbio’s E2 TPD technology. Last year, BPGbio acquired Berg and its AI platform.

Generate Biomedicines

In September, Generate Biomedicines agreed to apply its namesake Generate Platform in a potentially more than $1 billion collaboration with Novartis to discover and develop protein therapeutics for multiple unspecified disease areas. Generate has also been partnering with Amgen to discover and create protein therapeutics across several therapeutic areas and multiple modalities, including monoclonal and bispecific antibody drugs.

Generate emerged from stealth in 2020 when it completed a Series A totaling $50 million, funded solely by Flagship. A year later, Flagship led the company’s $370 million Series B financing, joined by several institutional co-investors. Earlier this year, Generate landed on CNBC’s “Disruptor 50” list of private companies “upending the classic definition of disruption.”

Insilico Medicine

Insilico Medicine’s lead candidate, ISM001-055 for idiopathic pulmonary fibrosis, performed well in a Phase IIa trial, showing improved forced vital capacity at 12 weeks, plus favorable safety and pharmacokinetics profiles. Insilico’s pipeline includes 31 disclosed programs, 9 with IND approvals.

Insilico intends to expand its pipeline and upgrade its Pharma.AI platform through a $100 million revolving loan facility with HSBC signed in November. Earlier this year, Insilico relocated its headquarters to Cambridge, MA, from New York City and Hong Kong, citing the Boston region’s large talent pool and critical mass of potential business development partners.

Model Medicines

Model Medicines has one of the largest disclosed pipelines of any AI-based drug discovery company. It has 192 compounds aimed at 26 therapeutic targets. All the compounds have been discovered through the company’s GALILEO platform, which is designed to investigate “constellations” of interacting atoms within 3D protein structures.

In April, a team of researchers from Model and partners posted a preprint identifying the RdRp Thumb-1 site, which represents a potentially druggable family of targets across positive-sense, single-stranded RNA viruses—all of which may be targetable by a single broad-spectrum antiviral drug.3 Model also disclosed a Phase I candidate, MDL-001, as a potential broad-spectrum nonnucleoside antiviral drug candidate. Also in Phase I is MDL-4101, which is designed to treat thyroid cancer by targeting bromodomain-containing protein 4.

Nimbus Therapeutics

Nimbus Therapeutics’ pipeline focuses on oncology, immunology, metabolic disorders, and other indications. In 2022, Takeda committed up to $6 billion to acquire Nimbus’ immunology candidate NDI-034858 (since renamed zasocitinib or TAK-279), a tyrosine kinase 2 inhibitor now in Phase II/III trials. In a recent JAMA Dermatology article, Phase IIb trial results for zasocitinib were presented by researchers who stated that the drug candidate “represents a potential new oral treatment for psoriasis.”

In 2016, Gilead Sciences committed up to $1.2 billion to acquire Nimbus’ NDI-010976, an allosteric acetyl-coenzyme A carboxylase inhibitor designed to treat metabolic dysfunction–associated steatohepatitis and potentially hepatocellular carcinoma and other diseases. Since 2022, Nimbus has partnered with Eli Lilly in an up-to-$496 million-plus collaboration to develop metabolic drugs activating a specific isoform of adenosine monophosphate–activated protein kinase.

Recursion Pharmaceuticals

Recursion Pharmaceuticals recently reported progress toward its goal of developing at least 100 pipeline candidates in roughly a decade, up from the current 7 disclosed candidates and “more than a dozen additional early discovery and research programs.” Most are expected to emerge internally. Recursion also hopes to grow its pipeline through its acquisition of Exscientia.

The company has dosed the first patient in a Phase II trial assessing REC-3964—the first new chemical entity to be developed through its RecurionOS Operating System—in recurrent Clostridioides difficile infection. Recursion has also announced positive topline Phase II data for its symptomatic cerebral cavernous malformation candidate, REC-994.

Relay Therapeutics

Recursion Therapeutics plans to launch a pivotal trial of its RLY-2608 as a second-line treatment for breast cancer in 2025, based on data showing that the drug plus fulvestrant led to clinically meaningful progression-free survival at the recommended Phase II dose of 600 mg twice daily in heavily pretreated patients with PI3Kα-mutated, HR-positive, and HER2-negative metastatic breast cancer. President and CEO Sanjiv Patel, MD, declared that Relay expects to fully fund the trial through to topline readout using its existing cash on hand. The company reported approximately $840 million in cash, cash equivalents, and investments at the end of the third quarter.

 

Schrödinger

Through a $10 million grant from the Bill and Melinda Gates Foundation, Schrödinger in August began expanding its computational platform to predict toxicology risk early in drug discovery. In October, Schrödinger recently presented data showing that in preclinical models, treatment with the company’s Wee1/Myt1 inhibitor SGR-3515 resulted in synergistic antitumor activity that led to deeper and more durable responses compared to inhibitors that target only Wee1 or Myt1. SGR-3515 is one of eight disclosed proprietary programs at Schrödinger. The company has 19 programs it is pursuing with partners, whichinclude Bristol Myers Squibb, Eli Lilly, Gilead, and Takeda.

 

 

References

  1. Narain NR. AI Can Give Pharma More Shots on Goal, But Robust Data and Validation Are Key to Drug Approvals. Forbes. Posted October 14, 2024. Accessed: November 11, 2024.
  2. Deep Pharma Intelligence. Artificial Intelligence for Drug Discovery: Landscape Overview Q1 2023. Accessed November 11, 2024.
  3. Woods V, Umansky T, Ramesh N, et al. Discovery of RdRp Thumb-1 as a novel broad-spectrum antiviral family of targets and MDL-001 as a potent broad-spectrum inhibitor thereof – Part I: A Bioinformatics and Deep Learning Approach. bioRxiv. Posted April 3, 2024. Accessed November 11, 2024.
  4. Armstrong AW, Gooderham M, Lynde C, et al. Tyrosine Kinase 2 Inhibition with Zasocitinib (TAK-279) in Psoriasis: A Randomized Clinical Trial. JAMA Dermatol. 2024; 160(10): 1066–1074.
Previous articleSix Companies Worthy of Note in AI Drug Discovery