A Gene Correlation Measurement Method for Spatial Transcriptome Data Based on Partitioning and Distribution
Authors:
- 1 August 2023
- journal article
- research article
- Published by SAGE Publications in Journal of Computational Biology
- Vol. 30 (8) , 877-888
Abstract
Spatial transcriptome (ST) technology provides both the spatial location and transcriptional profile of spots, as well as tissue images. ST data can be utilized to construct gene regulatory networks, which can help identify gene modules that facilitate the understanding of biological processes such as cell communication. Correlation measurement is the core basis for constructing a gene regulatory network. However, due to the high noise and sparsity in ST data, common correlation measurement methods such as the Pearson correlation coefficient (PCC) and Spearman correlation coefficient (SPCC) are not suitable. In this work, a new gene correlation measurement method called STgcor is proposed. STgcor defines vertexes as spots in a two-dimensional coordinate plane consisting of axes X and Y from the gene pair (X and Y). The joint probability density of Gaussian distribution of the gene pair (X and Y) is calculated to identify and eliminate outliers. To overcome sparsity, the degree, trend, and location of the distribution of vertexes are used to measure the correlation between gene pairs (X, Y). To validate the performance of the STgcor method, it is compared with the PCC and SPCC in a weighted coexpression network analysis method using two ST datasets of breast cancer and prostate cancer. The gene modules identified by these methods are then compared and analyzed. The results show that the STgcor method detects some special gene modules and cancer-related pathways that cannot be detected by the other two methods.Keywords
This publication has 32 references indexed in Scilit:
- Nanopore sequencing technology and tools for genome assembly: computational analysis of the current state, bottlenecks and future directionsBriefings in Bioinformatics, 2018
- SCANPY: large-scale single-cell gene expression data analysisGenome Biology, 2018
- Visualization and analysis of gene expression in tissue sections by spatial transcriptomicsScience, 2016
- Ancient DNA and the rewriting of human history: be sparing with Occam’s razorGenome Biology, 2016
- Predicting Hub Genes Associated with Cervical Cancer through Gene Co-Expression NetworksIEEE/ACM Transactions on Computational Biology and Bioinformatics, 2015
- Spatially resolved transcriptomics and beyondNature Reviews Genetics, 2014
- Unbiased classification of sensory neuron types by large-scale single-cell RNA sequencingNature Neuroscience, 2014
- Beyond modules and hubs: the potential of gene coexpression networks for investigating molecular mechanisms of complex brain disordersGenes, Brain and Behavior, 2013
- Two gene co-expression modules differentiate psychotics and controlsMolecular Psychiatry, 2012
- WGCNA: an R package for weighted correlation network analysisBMC Bioinformatics, 2008