Method and system for screening biomarkers of chronic atrophic gastritis based on multi-modal differential maps

Abstract

The present application belongs to the technical field of bioinformatics, and specifically relates to a chronic atrophic gastritis biomarker screening method and system based on a multi-modal differential network. The present application comprises: preprocessing metabolome data and microbiome data, respectively constructing intra-modal interaction graphs and inter-modal interaction graphs, and obtaining a multi-modal differential network by making a difference between the interaction graphs under different clinical groupings; based on the constructed differential network, a multi-level difference evaluation index is constructed: CDI is calculated at the node level to quantify the network topology and biological function differences of the biomarkers; at the structure level, a group specificity index GSI is defined to characterize the group specificity of the differential network structure; the feature ranking results obtained by CDI and GSI are fused and screened to obtain a biomarker set for atrophic gastritis classification prediction. The present application improves the effectiveness of biomarker screening under the condition of small sample and high dimension, and provides support for subsequent targeted cultivation of candidate colonies and metabolites and development of potential intervention target points.

Images