Panchamahabhuta Genomics: Tridosha disease prediction model

Tools

Item Type:	Dataset
Title:	Panchamahabhuta Genomics: Tridosha disease prediction model
Creators Name:	Pande, Amit
Abstract:	Trained Random Forest model for predicting disease category from protein Tridosha composition. Achieves 91.2% accuracy across 13 disease categories using 68,573 ClinVar pathogenic variants. Includes protein Prakriti database and feature encoders. Part of the Panchamahabhuta Genomics framework described in Pande et al. (2025).
Source:	Zenodo
Publisher:	CERN
Date:	10 February 2026
Official Publication:	https://doi.org/10.5281/zenodo.18053685

Repository Staff Only: item control page