Can AI predict the taste of medicine? Here’s what the study says…

 A dedicated team of researchers from the UCL Global Business School for Health and the UCL School of Pharmacy is on a mission to make medicines taste better, and they’re turning to artificial intelligence for assistance. Their goal is to overcome one of the most significant challenges in pediatric healthcare – improving the palatability of medications.

For children, the taste of medicine can be a major roadblock, especially when it comes to long-term treatments like HIV antiretrovirals and antibiotics for tuberculosis. This distaste often results in difficulties with treatment adherence, which can lead to treatment failure and complications while also contributing to the growing problem of antimicrobial resistance (AMR).

The team at University College London is now working diligently to make medicines more kid-friendly, and they’re doing it at a pace never before seen. They’re harnessing data collected from an intriguing device known as an “electric tongue” to develop an AI model capable of predicting the taste of medicines.

This cutting-edge model dissects a drug into various chemical descriptors, which are the key factors that determine its taste. The AI model then learns to connect these descriptors with the predicted levels of bitterness. Dr Hend Abdelhakim (, an assistant professor at the UCL Global Business School for Health, elaborates on the process: “We put the AI to work using a machine learning algorithm to essentially decipher the chemical and molecular structures and other physical characteristics that make a drug taste bitter. We’re exploring if there’s a clear relationship between these factors.”

But the AI doesn’t stop at bitterness

It can also detect other taste characteristics such as salty, sweet, sour, umami, and astringent. However, their primary focus remains on bitterness and astringency, as these are the tastes most likely to deter patients, especially children, from sticking to their medication regimens.

The potential impact of this AI is substantial. It promises to significantly expedite the drug development process by eliminating the need for early-stage human taste trials, which are not only time-consuming but also costly.

What’s truly exciting is that this AI model, once fully developed, will be made openly accessible. This means that pharmaceutical developers worldwide can tap into the valuable data on the taste profiles of commonly used drugs.

Taste is a critical factor in ensuring that people consistently take their medications

It’s an essential aspect of drug development. For instance, children often struggle with taking medicine due to taste, but this challenge isn’t limited to them. Adults, especially those on long-term medications like individuals with HIV, also face this issue.

Dr Abdelhakim and her team have taken a unique approach by utilising an electronic tongue to assess the bitterness of medicines and estimate how aversive the clinical dose may be. The electronic tongue measures how bitter molecules adhere to a plastic sensor that mimics the human tongue and then compares it with a clear sample to calculate the theoretical bitterness level of a medicine.

This innovative approach not only speeds up the testing process but also ensures accuracy. The researchers are collaborating with machine learning experts, including Dr David Shorthouse from the UCL School of Pharmacy, to further enhance drug development through the use of AI.

How it works

Using data from the electronic tongue, this AI model dissects a drug into a set of molecular descriptors, such as the number of atoms and the total surface area of the molecule. These descriptors are crucial in accurately predicting taste levels.

The development of this model is poised to benefit the entire pharmaceutical industry, as it will be an open-access tool. This means that pharmaceutical developers around the globe can leverage the wealth of data on medication palatability.

Dr Abdelhakim emphasises that the taste of medicines is a particular concern for children, who have a highly sensitive palate. This issue is especially critical for long-term diseases like HIV, where antiretroviral medications often have an unpleasant taste. Ensuring that young patients adhere to their medication regimens is essential for their well-being.

The importance of adherence to antibiotics cannot be overstated, particularly in the fight against antimicrobial resistance (AMR). Incomplete antibiotic courses can contribute to this global health threat, making it crucial to address this issue. Dr Abdelhakim underscores the significance, stating, “It’s a bigger problem for the rest of us.”