MDSVDNV: predicting microbe-drug associations by singular value decomposition and Node2vec.

Area of Science:

• Microbiology
• Pharmacology
• Bioinformatics

Background:

• Microbes play a vital role in human health, and imbalances in the microbiome can lead to disease.
• Identifying drug-associated microbes is crucial for advancing clinical medicine and understanding disease pathology.

Purpose of the Study:

• To propose a novel prediction model, MDSVDNV, for inferring potential microbe-drug associations.
• To leverage network embedding and matrix decomposition techniques for enhanced prediction accuracy.

Main Methods:

• The MDSVDNV model utilizes the Node2vec network embedding approach for linear representations.
• Singular value decomposition (SVD) matrix decomposition is employed for non-linear representations of microbe interactions.

Main Results:

• MDSVDNV achieved a high Area Under the Curve (AUC) value of 98.51% under 5-fold cross-validation.
• Experimental results demonstrate that MDSVDNV outperforms existing state-of-the-art methods in predicting microbe-drug associations.

Conclusions:

• MDSVDNV is an effective method for discovering latent microbe-drug associations.
• The model shows significant potential for future applications in clinical medicine and microbiome research.