Ensuring drug safety using AI models for adverse drug reaction prediction

An AI model developed to predict adverse drug reactions could potentially support early-stage drug safety assessment before clinical trials.

Adverse drug reactions (ADRs) are a significant cause of hospital admissions and treatment discontinuation worldwide. Conventional approaches often fail to detect rare or delayed effects of medicinal products. In order to improve early detection, a research team from the Medical University of Sofia developed a deep learning model to predict the likelihood of ADRs based solely on a drug’s chemical structure.

The model was built using a neural network trained using reference pharmacovigilance data. Input features were derived from SMILES codes – a standard format representing molecular structure. Predictions were generated for six major ADRs: hepatotoxicity, nephrotoxicity, cardiotoxicity, neurotoxicity, hypertension, and photosensitivity.

A flowchart illustrating chemical transformation, featuring molecular structures, fragment analysis, and decomposition stages.
Visual representation of SMILES and the process of molecular deconstruction. Adapted from Wu JN, Wang T, Chen Y, Tang LJ, Wu HL, Yu RQ. t-SMILES: a fragment-based molecular representation framework for de novo ligand design. Nat Commun. 2024 Jun 11;15(1): 4993. https://doi.org/10.1038/s41467-024-49388-6.

“We could conclude that it successfully identified many expected reactions while producing relatively few false positives,” the researchers write in their paper published in the journal Pharmacia, concluding it “demonstrates acceptable accuracy in predicting ADRs.”

Infographic detailing an AI model predicting adverse drug reactions for various compounds.

Testing of the model with well-characterized drugs resulted in predictions consistent with known side-effect profiles. For example, it estimated a 94.06% probability of hepatotoxicity for erythromycin, 88.44% for nephrotoxicity and 75.8% for hypertension in cisplatin. Additionally, 22% photosensitivity was predicted for cisplatin, while 64.8% photosensitivity was estimated for the experimental compound ezeprogind. For enadoline, a novel molecule, the model returned low probability scores across all ADRs, suggesting minimal risk.

Notably, these results demonstrate the model’s potential as a decision-support tool in early-phase drug discovery and regulatory safety monitoring. The authors acknowledge that performance of the infrastructure could be further enhanced by incorporating factors such as dose levels and patient-specific parameters.

Research article:

Ruseva V, Dobrev S, Getova-Kolarova V, Peneva A, Getov I, Dimitrova M, Petkova V (2025) In situ development of an artificial intelligence (AI) model for early detection of adverse drug reactions (ADRs) to ensure drug safety. Pharmacia 72: 1–8. https://doi.org/10.3897/pharmacia.72.e160997

Artificial neural networks could power up curation of natural history collections

Deep learning techniques manage to differentiate between similar plant families with up to 99 percent accuracy, Smithsonian researchers reveal

Millions, if not billions, of specimens reside in the world’s natural history collections, but most of these have not been carefully studied, or even looked at, in decades. While containing critical data for many scientific endeavors, most objects are quietly sitting in their own little cabinets of curiosity.

Thus, mass digitization of natural history collections has become a major goal at museums around the world. Having brought together numerous biologists, curators, volunteers and citizens scientists, such initiatives have already generated large datasets from these collections and provided unprecedented insight.

Now, a study, recently published in the open access Biodiversity Data Journal, suggests that the latest advances in both digitization and machine learning might together be able to assist museum curators in their efforts to care for and learn from this incredible global resource.

A team of researchers from the Smithsonian Department of BotanyData Science Lab, and Digitization Program Office recently collaborated with NVIDIA to carry out a pilot project using deep learning approaches to dig into digitized herbarium specimens.

Smithsonian researchers classifying digitized herbarium sheets.
Smithsonian researchers classifying digitized herbarium sheets.

Their study is among the first to describe the use of deep learning methods to enhance our understanding of digitized collection samples. It is also the first to demonstrate that a deep convolutional neural network–a computing system modelled after the neuron activity in animal brains that can basically learn on its own–can effectively differentiate between similar plants with an amazing accuracy of nearly 100%.

In the paper, the scientists describe two different neural networks that they trained to perform tasks on the digitized portion (currently 1.2 million specimens) of the United States National Herbarium.

The team first trained a net to automatically recognize herbarium sheets that had been stained with mercury crystals, since mercury was commonly used by some early collectors to protect the plant collections from insect damage. The second net was trained to discriminate between two families of plants that share a strikingly similar superficial appearance.

Sample herbarium specimen image of stained clubmoss
Sample herbarium specimen image of stained clubmoss.

The trained neural nets performed with 90% and 96% accuracy respectively (or 94% and 99% if the most challenging specimens were discarded), confirming that deep learning is a useful and important technology for the future analysis of digitized museum collections.

“The results can be leveraged both to improve curation and unlock new avenues of research,” conclude the scientists.

“This research paper is a wonderful proof of concept. We now know that we can apply machine learning to digitized natural history specimens to solve curatorial and identification problems. The future will be using these tools combined with large shared data sets to test fundamental hypotheses about the evolution and distribution of plants and animals,” says Dr. Laurence J. Dorr, Chair of the Smithsonian Department of Botany.

 

###

Original source:

Schuettpelz E, Frandsen P, Dikow R, Brown A, Orli S, Peters M, Metallo A, Funk V, Dorr L (2017) Applications of deep convolutional neural networks to digitized natural history collections. Biodiversity Data Journal 5: e21139. https://doi.org/10.3897/BDJ.5.e21139