JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

An oral microbiota-based deep neural network model for risk stratification and prognosis prediction in gastric cancer

  • 0Integrative Microecology Clinical Center, Shenzhen Clinical Research Center for Digestive Disease, Shenzhen Technology Research Center of Gut Microbiota Transplantation, The Clinical Innovation & Research Center, Shenzhen Key Laboratory of Viral Oncology, Department of Clinical Nutrition, Shenzhen Hospital, Southern Medical University, Shenzhen, China.
Journal of Oral Microbiology +

|

January 22, 2025

|

January 22, 2025

View abstract on PubMed

Summary

This summary is machine-generated.

This study identifies oral bacteria linked to gastric cancer (GC) survival. A Deep Neural Network (DNN) model using these markers accurately predicts GC prognosis, outperforming traditional staging methods.

Area Of Science

  • Microbiome research
  • Oncology
  • Bioinformatics

Background

  • Gastric cancer (GC) remains a significant global health challenge.
  • Accurate risk stratification and prognosis prediction are crucial for effective GC management.
  • The role of the oral microbiome in GC pathogenesis and progression is an emerging area of investigation.

Purpose Of The Study

  • To develop an oral microbiota-based model for gastric cancer (GC) risk stratification.
  • To predict GC prognosis using identified oral microbial markers.
  • To evaluate the performance of machine learning models in GC prognosis prediction.

Main Methods

  • Identification of oral microbial markers associated with GC prognosis in 99 GC patients.
  • Validation of marker predictive potential on an external cohort of 111 GC patients.
  • Construction and evaluation of Deep Neural Network (DNN), Random Forest (RF), and Support Vector Machine (SVM) models for GC prognosis prediction.

Main Results

  • Specific oral bacteria, including *Aggregatibacter*, *Filifactor*, *Moryella*, and *Leptotrichia*, were identified as significant prognostic markers for GC.
  • A DNN model based on these bacterial markers achieved high accuracy in predicting 3-year (AUC 0.814) and 5-year (AUC 0.912) survival.
  • The DNN model demonstrated superior performance compared to TNM staging, SVM, and RF models, with findings reinforced by external validation.

Conclusions

  • An oral microbiota-based DNN model shows promise for advancing GC prognosis prediction.
  • The identified oral bacterial markers may serve as novel biomarkers for GC.
  • Further research into the biological functions of these oral bacteria in GC progression is warranted.

Keywords: