Spatial Transcriptomic Analysis Using R-Based Computational Machine Learning Reveals the Genetic Profile of Yang or Yin Deficiency Syndrome in Chinese Medicine Theory

Abstract

Objectives. Yang and Yin are two main concepts responsible for harmonious balance reflecting health conditions based on Chinese medicine theory. Of note, deficiency of either Yang or Yin is associated with disease susceptibility. In this study, we aim to clarify the molecular feature of Yang and Yin deficiency by reanalyzing a transcriptomic data set retrieved from the GEO database using R-based machine learning analyses, which lays a foundation for medical diagnosis, prevention, and treatment of unbalanced Yang or Yin. Methods. Besides conventional methods for target mining, we took the advantage of spatial transcriptomic analysis using R-based machine learning approaches to elucidate molecular profiles of Yin and Yang deficiency by reanalyzing an RNA-Seq data set (GSE87474) in the GEO focusing on peripheral blood mononuclear cells (PBMCs). The add-on functions in R including GEOquery, DESeq2, WGCNA (target identification with a scale-free topological assumption), Scatterplot3d, Tidyverse, and UpsetR were used. For information in the selected GEO data set, PBMCs representing 20,740 expressed genes were collected from subjects with Yang or Yin deficiency (n = 12 each), based on Chinese medicine-related diagnostic criteria. Results. The symptomatic gene targets for Yang deficiency (KAT2B, NFKB2, CREBBP, GTF2H3) or Yin deficiency (JUNB, JUND, NGLY1, TNF, RAF1, PPP1R15A) were potentially discovered. CREBBP was identified as a shared key contributive gene regulating either the Yang or Yin deficiency group. The intrinsic molecular characteristics of these specific genes could link with clinical observations of Yang/Yin deficiency, in which Yang deficiency is associated with immune dysfunction tendency and energy deregulation, while Yin deficiency mainly contains oxidative stress, dysfunction of the immune system, and abnormal lipid/protein metabolism. Conclusion. Our study provides representative gene targets and modules for supporting clinical traits of Yang or Yin deficiency in Chinese medicine theory, which is beneficial for promoting the modernization of Chinese medicine theory. Besides, R-based machine learning approaches adopted in this study might be further applied for investigating the underlying genetic polymorphisms related to Chinese medicine theory.

Comprehensive Molecular Analyses of Notch Pathway-Related Genes to Predict Prognosis and Immunotherapy Response in Patients with Gastric Cancer

“…In the field of computer vision, PCA is a method that is frequently used for dimensionality reduction and feature extraction [ 31 ]. In order to study the potential differences that may exist between high-risk and low-risk groups, the “scatterplot3d” R tool was used [ 32 ].…”

Multi-component Chinese medicine formulas for drug discovery: State of the art and future perspectives

“…In modern Chinese medicine research, two novel methods-network pharmacology and integrative omics technology-are commonly integrated to investigate the therapeutic mechanisms of multi-component Chinese medicine formulas at a systematic investigation [117][118][119].…”

Optimizing a Interpretable Diagnostic Model for Colorectal Cancer Based on Yin Deficiency Pattern Characteristic Genes Using 21 Machine Learning Algorithms and Bayesian Opitimization

“…We selected the RNA dataset GSE87474 [19] from the GEO database [20] and chose 20 samples, including eight normal samples and 12 samples from patients with YDP. We used CRC RNA-seq gene expression data (n=434) from TCGA [21] for the analysis of CRC core biomarkers related to YDP.…”