Machine Learning Prediction of Antimicrobial Potential in Pleurotus ostreatus Cultivated on Cassava Peel: Dataset and Analysis Code

This repository contains the raw and processed antimicrobial assay data, mycochemical composition data, analysis scripts, and results supporting a short communication on machine learning prediction of antimicrobial potential in Pleurotus ostreatus cultivated on cassava peel. Antimicrobial activity (zone of inhibition and minimum inhibitory concentration) of ethanolic and aqueous P. ostreatus extracts was evaluated in triplicate against seven bacterial and fungal pathogens. A Random Forest algorithm with leave-one-out cross-validation (LOOCV) was used to predict zone of inhibition (regression) and classify minimum inhibitory concentration into high/low sensitivity classes. Contents:- data/raw/: raw triplicate data transcribed from laboratory notebook records- data/processed/: merged, analysis-ready dataset- scripts/: Python scripts to reproduce the full analysis pipeline (data merging, model training, figure generation)- results/: model outputs (LOOCV predictions, metrics, feature importances) and all manuscript figures See README.md and DATA_DICTIONARY.md within the archive for full details and reproduction instructions.