Nov 10, 2025 08:00:00

What will come from understanding 'off-target effects,' where drugs have effects that are different from those expected?

Off-target effects are effects caused by a drug acting on a target other than its intended target. Medical researcher Abhishaike Mahajan (@owl_posting) explains why understanding off-target effects is important.

Mapping the off-target effects of every FDA-approved drug in existence (EvE Bio)
https://www.owlposting.com/p/mapping-the-off-target-effects-of

Pharmaceutical companies conduct research that is optimized to determine whether a drug is effective for a specific condition. The indicators are simple, such as a reduction in specific biomarkers, improved mood, or improved function, with little attention paid to off-target effects. Pharmaceutical companies have limited time and resources, so it is inevitable that they limit their investments to whether a drug is effective or not, discarding everything else.

Naturally, if off-target effects are discovered during drug development, resources are allocated to address them. However, mapping every possible off-target effect—such as unintended receptor binding, deviations in pathways of action, or unintended gene inhibition—during clinical trials is labor-intensive and unprofitable. Therefore, off-target effects are typically investigated only if concerns emerge during post-marketing surveillance.

While there are many reasons why pharmaceutical companies don't research off-target effects, Mahajan lists three reasons why understanding them is important.

◆Redevelopment of existing drugs (drug repositioning, drug repurposing)

Drug redevelopment refers to the repurposing of an existing drug developed for a specific purpose for a different therapeutic purpose. A study of 197 drugs approved by the U.S. Food and Drug Administration (FDA) between 1997 and 2020 found that approximately 30% of FDA-approved drugs had new indications added after approval, and a significant number of drugs have been found to have effects beyond their original use.

While many clinicians intuitively explore the possibility of redeveloping existing drugs, a 2006 study found that 73% of off-label prescriptions had little or no scientific basis. However, Mahajan argues that even if redeveloping existing drugs effectively and efficaciously is difficult, it is well worth the effort.

A major benefit of redeveloping existing drugs is the overwhelmingly lower speed and cost required for approval compared to developing new drugs. Developing new drugs requires not only research into various drug candidates, but also the collection of toxicity data already collected and repeated clinical trials. On the other hand, because existing drugs have already passed various clinical trials and safety tests, costs can be reduced by billions of dollars (hundreds of billions of yen), and approval can be achieved in just a few years.

However, while new drugs come with new patents and full exclusivity, redevelopment of existing drugs typically targets drugs that are off patent or about to be on patent expiry, so even if a new use is approved, the profits for the company that discovered that use are not as large.

◆ Validation data for machine learning models
While the use of machine learning models has been increasing in drug discovery in recent years, publicly available datasets have issues with noise and experimental methods, so model developers are essentially forced to develop models using in-house datasets. Furthermore, in-house datasets often have limited scope, are biased toward specific molecular classes, or lack reproducibility.

To make matters worse, this data is often cherry-picked around success stories or focused on well-studied targets, introducing bias that limits generalizability. Therefore, unbiased, third-party generated datasets of off-target effects can be used as validation data for machine learning models.

◆Multiple pharmacology

Polypharmacology refers to the phenomenon where a specific compound has activity against multiple targets. The phenomenon is not much different from off-target effects, but in recent years, drugs that intentionally target multiple targets are being developed to achieve more effective action with a single drug.

For example, Ozempic (semaglutide), a type 2 diabetes medication, targets the GLP-1 receptor, which suppresses appetite and slows digestion, while Manjaro (tirzepatide) , which was released later, acts not only on the GLP-1 receptor but also on GIP , which promotes insulin response, resulting in greater weight loss.

By enabling a single drug to act on multiple targets, researchers can also avoid polypharmacy, which involves combining multiple drugs to manage a specific condition. While polypharmacy can be useful for managing conditions that cannot be treated with a single drug alone, it can also have adverse effects on the patient's physiology. Furthermore, the more drugs a patient is given, the more difficult it becomes to predict their overall interactions, increasing the risk of unexpected outcomes.

There is no empirical evidence that off-target effect datasets are useful for polypharmacology studies, but Mahajan believes that off-target effect datasets may provide a missing link for rational polypharmacology studies.

A non-profit research organization called EvE Bio is already working on building a dataset called 'EvE Bio Data,' which maps the off-target effects of all FDA-approved drugs. EvE Bio Data is released under a non-profit Creative Commons license, making it free for researchers to use and available for licensing by commercial organizations. Surprisingly, the most popular market demand is for validation data for machine learning models, and several commercial organizations are already in discussions to use it for this purpose.

EvE Bio Data
https://data.evebio.org/