An Investigation into the Factors Governing Drug Absorption and Food Effect Prediction Based on Data Mining Methodology

Machine learning predicts the effect of food on orally administered medicines

“…However, only LR was used, a basic statistical technique, which can only handle linear relationships between input and output data, whereas the modern ML approaches used here can handle more complex relationships between inputs and outputs. A further study used a RF for food effect prediction and reported with a moderate Kappa (a metric comparing the observed accuracy with the expected accuracy) from the modelling of a dataset of 53 drugs and 11 drug properties (Gatarić and Parojčić, 2019). The most recent study used artificial neural networks (ANNs) and SVM to predict the food effect using a dataset of 141 drug compounds brought to market in the last 5 years.…”

“…Our study aimed to use ML technologies to predict the food effect from an extensive database of over 300 drugs with a diverse set of chemical features and over 20 drug properties or features. Previous studies have investigated ML in the prediction of the food effect on smaller datasets with different methodologies (Bennett-Lenane et al, 2021;Gatarić and Parojčić, 2019;Gu et al, 2007). As opposed to a 'plug-and-play' standard ML approach, a strategic approach was implemented.…”