Oral_voting_transfer: classification of oral microorganisms' function proteins with voting transfer model.

Area of Science:

• Microbiology
• Bioinformatics
• Computational Biology

Background:

• The oral microbial ecosystem is complex and linked to various human diseases, including inflammation, cancer, diabetes, and neurological disorders.
• Oral microbial proteins play significant roles in disease pathogenesis, making their study critical for understanding disease progression.
• Traditional experimental methods struggle with the high-dimensional and imbalanced data characteristic of oral microbial research.

Purpose of the Study:

• To propose and evaluate a novel computational method, oral_voting_transfer, for classifying oral microorganisms.
• To address the challenges of high-dimensional and imbalanced data in oral microbial classification.
• To facilitate a deeper understanding of the role of oral microbes in human health and disease.

Main Methods:

• Developed a novel classification method named oral_voting_transfer.
• Utilized three key features for the classification of five specific oral microorganisms: *Streptococcus mutans*, *Staphylococcus aureus*, abiotrophy adjacent, bifidobacterial, and *Capnocytophaga*.
• Employed a transfer learning approach, adapting a model successful in organelle protein classification, combined with local classifiers and a voting mechanism.

Main Results:

• The oral_voting_transfer method demonstrated well-performed classification of the five targeted oral microorganisms.
• The developed tool is standalone and provides a robust solution for oral microbial classification tasks.
• Source code for the oral_voting_transfer method is publicly available for research use.

Conclusions:

• The oral_voting_transfer method offers an effective solution for classifying complex oral microbial data.
• This approach can aid in analyzing the progression of oral microbiome-associated diseases.
• The availability of the tool promotes further research in oral microbiology and its clinical implications.