The importance of artificial intelligence and machine learning (AI/ML) has not been lost on drug development companies. Recently, to help accelerate the discovery of therapies to treat COVID-19, several deals have been established to help deploy those tools. For example, Abcellera Biologics Inc., of Vancouver, British Columbia, and Eli Lilly and Co., of Indianapolis, agreed to co-develop antibody products for treating and preventing COVID-19. The collaboration will build on Abcellera's pandemic response platform, developed under the DARPA Pandemic Prevention Platform (P3) program, and Lilly's global capabilities for rapid development, manufacturing and distribution of therapeutic antibodies.

Within one week of receiving a blood sample from one of the first U.S. patients who recovered from COVID-19, Abcellera screened more than 5 million immune cells looking for those that produced functional antibodies that helped the patient neutralize the virus and recover from the disease and identified more than 500 unique fully human antibody sequences. The next step in their work is to screen those antibodies to find the ones most effective in neutralizing SARS-CoV-2. The goal is to be testing potential new therapies in patients within the next four months.

Under the terms of the agreement, Abcellera and Lilly have committed to equally share initial development costs toward a product, after which Lilly will be responsible for all further development, manufacturing and distribution.

Intelligence to drugs

Another Canadian company, Cyclica Inc., of Toronto, which is leveraging artificial intelligence and computational biophysics, and Beijing’s Institute of Materia Medica, Chinese Academy of Medical Sciences, are collaborating to discover antiviral drug candidates for COVID-19 and to explore opportunities to design multitargeted antiviral compounds. The institute will conduct in vitro and in vivo antiviral assessment for molecules proposed by Cyclica. The two will also engage in a longer-term effort to design multitarget antiviral compounds to reduce drug resistance.

Fargo, N.D.-based Immunoprecise Antibodies Ltd. (IPA), through its subsidiary, Talem Therapeutics, and the New York-based company Evqlv Inc., announced a new collaboration aimed at accelerating the global effort to develop lead therapeutic and vaccine candidates against the SARS-CoV-2 virus. Evqlv, an artificial intelligence company focused on accelerating the discovery and optimization of antibody therapies, will perform in silico antibody design and discovery to generate optimized antibody sequences leveraging lead candidate data from IPA's discovery technologies, as well as previously published sequences.

Therapeutic candidates developed through that collaboration will remain the ownership of IPA, with commercial royalties issued to Evqlv upon completion of predefined, commercial milestones.

AI Business development robust

In addition to deals targeting COVID-19, the first quarter has seen a number of deals that link AI with drug discovery. Among them, South Korea's Bridge Biotherapeutics Inc. has enlisted San Francisco-based AI drug discovery shop Atomwise Inc. to help it identify up to 13 small-molecule inhibitors of Pellino proteins and other targets in an effort to develop new therapies for inflammatory diseases. Atomwise stands to receive an up-front payment of undisclosed value, as well as milestone and royalty payments under terms of the deal. Presuming success, it estimates the partnership's payoff could reach up to $1.08 billion.

Pellino E3 ubiquitin ligases are a target area Seongnam-based Bridge has pioneered with BBT-401, the first-in-class Pellino-1 peptide inhibitor, currently in phase II as a potential treatment for ulcerative colitis. The new partnership with Atomwise will lean on the AI expert’s structure-based discovery technology, Atomnet, to help pare millions of potential molecules down to a curated subset ready for preclinical testing.

Pfizer Inc., of New York, will use Hong Kong-based Insilico Medicine Inc.’s machine learning technology and platform to identify therapeutic targets for a variety of diseases. Insilico's platform imagines potential protein structures based on existing research and certain pre-programmed design criteria.

Market turbulence

Although investors warmed to companies involved in this field during 2019, the BioWorld Artificial Intelligence price-weighted index, which includes biopharmaceutical companies, medical devices and health care services companies that are utilizing AI, the market turbulence and flight from equities has taken a toll on member companies in the first quarter. As a result, the index is currently down 27.5% in value since the beginning of the year.

Leading group decliner so far this month is Bioxcel Therapeutics Inc., of New Haven, Conn., whose shares (NASDAQ:BTAI) have dipped 60%, erasing nearly all the gains it enjoyed earlier in the year. The biopharmaceutical company is utilizing artificial intelligence to identify improved therapies in neuroscience and immuno-oncology.

In February, it raised $60 million in net proceeds from a common stock offering. The funds will support its clinical development programs that include BXCL-501, an investigational sublingual thin film of dexmedetomidine, a selective alpha-2A adrenergic receptor agonist, designed for the treatment of acute agitation. Pivotal phase III trials for the acute treatment of agitation in patients with schizophrenia (Serenity I) and bipolar disorder (Serenity II) are underway, and top-line data readouts from both trials are expected midyear.