GBDTSVM: Combined Support Vector Machine and Gradient Boosting Decision Tree Framework for efficient snoRNA-disease...

Area of Science:

• Biochemistry and Molecular Biology
• Computational Biology
• Genomics

Background:

• Small nucleolar RNAs (snoRNAs) play a crucial role in human disease pathogenesis.
• Accurate identification of snoRNA-disease associations (SDAs) is vital for disease progression understanding and therapeutic development.
• Traditional experimental methods for SDA identification are resource-intensive, necessitating efficient computational approaches.

Purpose of the Study:

• To develop a novel and efficient machine learning model, GBDTSVM, for predicting snoRNA-disease associations.
• To leverage Gradient Boosting Decision Tree (GBDT) and Support Vector Machine (SVM) for feature extraction and classification.
• To enhance prediction accuracy using Gaussian integrated profile kernel similarity.

Main Methods:

• The GBDTSVM model integrates GBDT for feature representation and SVM for classification of SDAs.
• Gaussian integrated profile kernel similarity is applied to both snoRNAs and diseases to improve prediction accuracy.
• The model's performance is evaluated on multiple datasets (MDRF, LSGT, PsnoD).

Main Results:

• GBDTSVM achieved superior performance compared to existing state-of-the-art methods.
• The model demonstrated high predictive accuracy with an AUROC of 0.96 and AUPRC of 0.95 on the MDRF dataset.
• A case study validated the model's predictions for top-ranked snoRNAs across twelve prevalent diseases.

Conclusions:

• The GBDTSVM model presents a robust and efficient computational tool for predicting snoRNA-disease associations.
• This framework has the potential to significantly advance research in snoRNA-related diseases.
• The study provides open-source code and datasets for reproducibility and further research.