This repository contains the source codes and datasets for our paper "Semantic-Enhanced Heterogeneous Graph Learning for Identifying ncRNAs Associated with Drug Resistance". It includes the implementation of the proposed predictor, along with datasets and tutorials to help users reproduce and extend our work. *Heterogeneous graph: Files for constructing the heterogeneous graph. *Train-val-test-data: Training, validation, and test sub-datasets of ncRNA–drug resistance associations. *Extract latent feature: Embeddings of ncRNA–drug resistance pairs extracted from iNcRD-HG. *Case studies: Datasets used for case study experiments. *Main code: Source codes covering main experimental workflows and analyses in iNcRD-HG. *ChemBERTa-zinc-base-v1: Source codes for extracting drug features.