Dataset of Medicinal Plants with Immunomodulatory Properties: Phytotherapeutic Attributes and Binary Classification Labels

This dataset accompanies the manuscript "Unveiling the Immunomodulatory Potential of Medicinal Plants: A Machine Learning Framework coupled with Experimental Validation" by Assouab et al. (2026). It contains structured data on 183 medicinal plants curated from over 100 peer-reviewed articles investigating their immunomodulatory properties. Each entry includes the following attributes: botanical species name (SPECIES), geographical region (REGION), used plant part (USED PART), metabolite content (METABOLITE CONTENT), tested microorganism (TESTED MICROORGANISM), and effective concentration (EFFECTIVE CONCENTRATION). The binary target variable indicates whether induction (1) or inhibition (0) of iNOS and cytokine gene expression was observed. The dataset also includes coded versions of the features, a validation set of 4 plants with known immunomodulatory status (Kalanchoe pinnata, Phyllanthus amarus, Eremanthus crotonoides, Empetrum nigrum), and a prediction set of 11 Moroccan medicinal plants with uncertain immunomodulatory effects. The data were used to train and evaluate machine learning classifiers using pre-trained language model embeddings for immunomodulatory activity prediction.