It is a system that through data would be able to predict side effects or reactions in patients when receiving some type of medication.
Paraguayan researcher Diego Galeano, who is part of the National Investigator Inventive Program (PRONII) of the National Council of Science and Technology (CONACYT) in collaboration with machine learning and artificial intelligence (AI) specialist Alberto Paccanaro, has developed this AI model that predicts the likelihood of predicting reactions or side effects in patients when receiving their treatment.
In September the article: Predicting the Frequency of Drug Side Effectswas published in the journal Nature Communications. “We present a machine learning framework for computationally predicting frequencies of drug side effects. Our matrix decomposition algorithm learns latent signatures of drugs and side effects that are both reproducible and biologically interpretable. We show the usefulness of our approach on 759 structurally and therapeutically diverse drugs and 994 side effects from all human physiological systems. Our approach can be applied to any drug for which a small number of side effect frequencies have been identified, in order to predict the frequencies of further, yet unidentified, side effects.”
The purpose of this novel method is to predict the side effects of medications to avoid risks in patients that may become lethal. Even another problem I could solve is avoiding medicine shortage. In addition to the benefits to end consumers of drugs, this algorithm would save a huge amount of money on pharmaceutical companies' investments to develop new drugs.
Alberto Paccanaro, Professor of Computational Biology at Royal Holloway, University of London, highlighted the relevance of this project: “There are several examples of medicines that had to be withdrawn from the market due to unknown side effects. Our work will help develop safer medicines for patients.”
“Our approach for predicting the frequencies of drug side effects is to use a matrix decomposition algorithm that learns a small set of latent features (or signatures) that encode the biological interplay between drugs and side effects,” they explain in the Discussion section in the article.
The drugs were grouped according to their main anatomical, therapeutic and chemical classes, while side effects were grouped according to their organ and system classification categories in MedDRA (Medical Dictionary for Regulatory Activities).
The machine learning algorithm was designed so that pharmaceutical companies can create more accurate drugs and perhaps in less time, so it is sought to inform that industry first so that they know what are the most recurrent side effects in patients depending on the drugs.
The article published in Nature Communications is available for free at the following link: https://www.nature.com/articles/s41467-020-18305-y#Sec8
