Robust gene coexpression networks using signed distance correlation

Abstract

Motivation Even within well studied organisms, many genes lack useful functional annotations. One way to generate such functional information is to infer biological relationships between genes/proteins, using a network of gene coexpression data that includes functional annotations. However, the lack of trustworthy functional annotations can impede the validation of such networks. Hence, there is a need for a principled method to construct gene coexpression networks that capture biological information and are structurally stable even in the absence of functional information. Results We introduce the concept of signed distance correlation as a measure of dependency between two variables, and apply it to generate gene coexpression networks. Distance correlation offers a more intuitive approach to network construction than commonly used methods such as Pearson correlation and mutual information. We propose a framework to generate self-consistent networks using signed distance correlation purely from gene expression data, with no additional information. We analyse data from three different organisms to illustrate how networks generated with our method are more stable and capture more biological information compared to networks obtained from Pearson correlation or mutual information. Supplementary information Supplementary Information and code are available at Bioinformatics and https://github.com/javier-pardodiaz/sdcorGCN online.

Towards a global investigation of transcriptomic signatures through co-expression networks and pathway knowledge for the identification of disease mechanisms

“…For instance, Pardo-Diaz et al. ( 66 ) recently presented a novel method that adds directionality into the co-expression network. Finally, while we constructed a human interactome network from multiple pathway and interaction databases, the majority of proteins from the co-expression networks could not be mapped to the network, highlighting the incompleteness of the current interactome.…”

Signed Distance Correlation (SiDCo): A network analysis application of distance correlation for identifying metabolic networks disrupted in Dementia with Lewy Bodies

“…Distance correlation values with p>0.01 were set to zero keeping only the most statistically significant correlations. Sign of distance correlation was equated to the sign of Pearson correlation calculation as shown by 16 and implemented in-house as: . Calculations we performed using Matlab in-house routing.…”

“…Distance correlation was developed by Székely et al 15 , as a measure of dependence between variables including linear and non-linear correlations as well as correlation between vectors of different lengths. Additionally, applications of distance correlation in genomics have shown that although distance correlation calculation provides only positive values, it is possible to use sign measure from Pearson’s correlation for the determination of signed distance correlation measure 16 . In addition to correlation network development, methods from clustering can provide alternative approaches for interaction or dependence network determination.…”

COGENT: evaluating the consistency of gene co-expression networks

“…A full worked example can be found in Supplementary Section S5 and in the COGENT tutorial. COGENT has also been used to assess signed distance correlation as a measure of gene co-expression ( Pardo-Diaz et al , 2021 ). This application further shows that network construction methods prioritized by COGENT also capture more protein–protein interaction data than methods which were not prioritized.…”