Ranking differential genes in co-expression networks

Abstract

Identifying the genes that change between two conditions, such as normal versus cancer, is a crucial task in understanding the causes of diseases. Differential networking has emerged as a powerful approach to achieve this task and to detect the changes in the corresponding network structures. The goal of differential networking is to identify the differentially connected genes between two networks. However, the current differential networking methods primarily depend on pair-wise comparisons of the genes based on their degrees in the two networks. Therefore, these methods cannot capture all the topological changes in the network structure. In this paper, we propose a novel differential networking algorithm, DiffRank, to rank the genes based on their contribution to the differences between two gene co-expression networks. To achieve this goal, we define two novel scoring measures: a local structure measure, differential connectivity, and a global structure measure, differential betweenness centrality. These measures are combined within a PageRank-style framework and optimized by propagating them through the network. Finally, the genes are ranked based on the their propagated scores. We demonstrate the effectiveness of DiffRank on several gene expression datasets, and we show that our method provides biologically interesting rankings.

Sc-compReg enables the comparison of gene regulatory networks between conditions using single-cell data

“…We also compare our method with any combinations of two gene regulatory network inference methods and two differential regulatory detection methods. Based on benchmarking literature of gene regulatory network inference from single-cell data 16 and differential regulatory network detection 14 , we choose GENIE3 17 and PIDC 18 as methods for network inference, and choose diffK 19 and diffRank 20 as methods for network comparison. The results showed that our method outperforms all those methods (Fig.…”

From time-series transcriptomics to gene regulatory networks: A review on inference methods

“…For example, using single-cell ATAC-seq, patterns of gene expression can be combined with chromatin accessibility profiles, thus identifying cell subpopulations and groups of cells at different developmental stages [ 124 ]. Consequently, several differential network analysis algorithms have been developed, such as DiffRank [ 125 ], dcanr R package [ 126 ], DiffK [ 127 ], DINA [ 128 ], to name a few. We refer the reader to [ 129 ] for a statistical perspective of DiNA and [ 130 ] for a DiNA algorithm comparison.…”