Finding potential lncRNA-disease associations using a boosting-based ensemble learning model.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Long non-coding RNAs (lncRNAs) are crucial in cancer development and serve as potential prognostic biomarkers.
• Identifying lncRNA-disease associations (LDAs) is vital for biomarker discovery and therapeutic strategies.
• Experimental methods for LDA identification are time-consuming and expensive.

Purpose of the Study:

• To develop a novel computational framework, LDA-SABC, for accurate lncRNA-disease association prediction.
• To leverage machine learning for efficient and cost-effective LDA identification.
• To identify potential lncRNA biomarkers for lung cancer.

Main Methods:

• Developed LDA-SABC, a boosting-based framework integrating LightGBM and AdaBoost with convolutional neural networks.
• Utilized singular value decomposition (SVD) for extracting lncRNA-disease association features.
• Validated performance using five-fold cross-validation and compared against existing LDA inference methods.

Main Results:

• LDA-SABC demonstrated superior performance over four classical methods in precision, recall, accuracy, F1 score, AUC, and AUPR.
• The framework successfully predicted potential lncRNA biomarkers for lung cancer.
• Identified 7SK and HULC as potential biomarkers for non-small-cell lung cancer (NSCLC) and lung adenocarcinoma (LUAD), respectively.

Conclusions:

• LDA-SABC offers a robust and efficient computational approach for predicting lncRNA-disease associations.
• The identified lncRNAs (7SK, HULC) warrant further investigation as potential biomarkers for lung cancer subtypes.
• This method can significantly aid in advancing LDA identification and cancer biomarker discovery.