<img height="1" width="1" style="display:none;" alt="" src="https://px.ads.linkedin.com/collect/?pid=384506&amp;fmt=gif">
Skip to content
Sign Up
    February 21, 2023

    Drug Discovery Industry Roundup with Barry Bunin — February 21, 2023

    Barry Bunin, PhD Founder & CEO Collaborative Drug Discovery Barry Bunin, PhD
    Founder & CEO
    Collaborative Drug Discovery

    Drug Discovery Industry Roundup with Barry Bunin

    Can Generative AI Help Defeat Drug-Resistant Pathogens? Forbes carries an article on IBM's generative AI, noting that applied to drug discovery, as often claimed for new technology that it "has the potential to save millions of lives around the world." With these types of claims, I often apply the framework that although new technologies may be overhyped in the short-term, their real impact is often underappreciated until the long-term. As an example, the article points to antibiotic-resistant superbug bacteria. Drug-resistant diseases kill 700,000 people annually around the world; by 2050, that number is expected to rise to 10 million deaths per year. Dr. Payel Das, an IBM Master Inventor at IBM Research lead a collaboration that utilized AI to synthesize and evaluate 20 unique antimicrobial peptide designs, chosen from a pool of 90,000 sequences. "The AI models were specifically designed to combat antibiotic resistance, incorporating controls for broad-spectrum efficacy and low toxicity, and slowing down the emergence of resistance," Forbes reports. "The team tested these designs against a diverse range of gram-negative and gram-positive bacteria, which led to the identification of six successful drug candidates." At CDD we find these trends exciting. We have our own research programs exploring the real scope and limitations of generative AI, with ChatGPT as the current poster child, for aiding drug discovery efforts.

      * * *

    "A Game Changer for Weight Loss." That headline from The Week is the latest in a number of publications looking at the benefits-and cautions-of a new wave of anti-obesity drugs. The article captures the excitement: "For decades, researchers have chased a drug that could significantly reduce weight without dangerous side effects. Now they say that moment has arrived. A pair of drugs, Ozempic and Wegovy - different formulations of the compound semaglutide - have shown stunning effectiveness in helping people lose weight by suppressing their appetites." Potential side effects include loss of energy, nausea, diarrhea, racing heartbeat, and what one podcaster termed her "power puking." Doctors also worry that if a weekly shot can produce weight loss, people will dismiss the need to exercise and improve their diets. Another concern is that the drugs are already being used by people who are not obese but have seized on them as an easy way to shed 5 or 10 unwanted pounds. "Such worries, though, are outstripped by excitement about the drugs' transformative potential," the story reads. "Medicines that can help patients shed 30 to 50 pounds or more are a 'game changer,' said John Buse, an endocrinologist at the University of North Carolina medical school. 'Obesity is on the ropes.'" These are drugs we continue to monitor closely in "The Pharmaceutical KnoweldgeBase" which we recently launched at www.pharmaKB.com.

      * * *

    New Use for an Old Drug-Curbing Binge Drinking. The New York Times carries the intriguing headline: "Binge Drinking May be Curbed With a Pill," describing a 12-week study, in which men wanting to reduce their binge drinking but who weren't severely dependent on alcohol were given a pill for whenever they felt cravings or anticipated a period of heavy drinking. It found binge drinkers may benefit from taking a dose of the medication naltrexone before consuming alcohol, a finding that may be welcomed now that alcohol-related deaths in the United States have surpassed 140,000 a year. "Taking naltrexone on an as-needed basis rather than as a daily dose may be more tolerable for some people because it allows their dopamine levels to recover in between uses," the article reads. "The approach could also let people feel more in control of their treatment. The practice is more widely embraced in Europe, where regulators in 2013 approved the medication nalmefene for similarly targeted dosing by people trying to drink less alcohol.

      * * *

    Machine Learning Speeding Discovery of Reaction Mechanisms. Derek Lowe, in his blog for Science, has a post titled "Burrowing Into Reaction Data, Automatically," celebrating the use of machine learning to lend a hand in what can be a complex inquiry. As with all machine learning processes, the results are only as good (and as structured) as the data building the training sets.  Writing about a new paper that shows an impressive application of machine learning, he notes the authors looked at twenty common catalytic mechanisms in several categories, incorporating mechanisms featuring steps like catalyst dimerization or the interaction between two different catalyst species entirely, catalyst activation processes, and similarly, various catalyst deactivation ones as well. "These things can be described in their ideal cases by sets of differential equations, and the paper uses these to generate a large pile of simulated (but very realistic) kinetic data across a wide range of time points to feed into an ML system. Beats running the experiments! They used two varieties of neural networks, one to handle the regular time-point data and another to deal with nontemporal stuff like initial concentrations of catalysts, and so on. The resulting model does a really solid job of classifying mechanisms from raw kinetic data."  We have our own research projections looking at data annotation, with the patented Assay Annotation technology demonstrated at www.bioassayexpress.com on over three thousand MLPCN assays now being incorporated into an Assay Management capability within the CDD Vault Assay & Registration module this quarter.  A bit further down the road, we are working actively as part of the publicly announced NCATS ASPIRE project with MIT Professor Connor Coley's ORD (Open Reaction Database) schema to more fully annotate reactions within the CDD Vault ELN (Electronic Laboratory Notebook) - which is a foundational step for further, more sophisticated analyses.


    Barry A. Bunin, PhD, is the Founder & CEO of Collaborative Drug Discovery, which provides a modern approach to drug discovery research informatics trusted globally by thousands of leading researchers. The CDD Vault is a hosted biological and chemical database that securely manages your private and external data.

    Other posts you might be interested in

    View All Posts
    CDD Blog
    8 min   July 22, 2024
    Drug Discovery Industry Roundup with Barry Bunin — July, 22 2024
    Read More
    CDD Blog
    34 min   July 1, 2024
    Critical Data Considerations for All Lifecycle Stages of a Biotech Startup
    Read More
    CDD Vault Snack
    3 min   June 27, 2024
    Vault Snack #24 - New Interface, Features & Infrastructure for the CDD Vault ELN
    Read More