Genome-Wide Study of Colocalization between Genomic Stretches: A Method and Applications to the Regulation of Gene Expression

Abstract

In this paper, we describe a method for the study of colocalization effects between stretch–stretch and stretch–point genome tracks based on a set of indices varying within the (–1, +1) interval. The indices combine the distances between the centers of neighboring stretches and their lengths. The extreme boundaries of the interval correspond to the complete colocalization of the genome tracks or its complete absence. We also obtained the relevant criteria of statistical significance for such indices using the complete permutation test. The method is robust with respect to strongly inhomogeneous positioning and length distribution of the genome tracks. On the basis of this approach, we created command-line software, the Genome Track Colocalization Analyzer. The software was tested, compared with other available packages, and applied to particular problems related to gene expression. The package, Genome Track Colocalization Analyzer (GTCA), is freely available to the users. GTCA complements our previous software, the Genome Track Analyzer, intended for the search for pairwise correlations between point-like genome tracks (also freely available). The corresponding details are provided in Data Availability Statement at the end of the text.

Preferential Co-Expression and Colocalization of rDNA-Contacting Genes with LincRNAs Suggest Their Involvement in Shaping Inter-Chromosomal Interactions with Nucleoli

“…Therefore, we searched for supposed correlations between residing sites of rDNA-contacting genes and corresponding lincRNAs. Recently, a tool for the genome-wide study of colocalization between genomic stretches was described: the Genome Track Colocalization Analyzer (GTCA) [ 17 ].…”

“…This colocalization is represented by the lncRNAs’ centers residing within co-expressing genes at a distance of approximately one quarter of the gene length from the gene boundary ( IO = −1 if the centers of colocalizing stretches coincide, IO = 0 if the centers of the first set of stretches are at the boundaries of the second set, and IO = 1 if the centers of stretches are located at the maximum possible distance from each other; see Methods). The analysis was performed using the GTCA [ 17 ], as described in Methods.…”

“…The colocalization of 4C-rDNA-associated lincRNA co-expressing genes and lincRNAs was assessed using a recently developed tool for the genome-wide study of colocalization between genomic stretches: the Genome Track Colocalization Analyzer (GTCA) [ 17 ]. First, the nearest neighbors were determined from the positions of the stretch centers.…”

“…We applied the complete permutation test to avoid uncertainty in sampling. The significance of deviation between the corresponding mean values is and can be assessed using the following normalization, since the resulting statistics proved to be close to Gaussian [ 17 ]: where …”

“…Statistical simulations [ 17 ] revealed that ζ can be transformed to a standard Gaussian statistic, with a mean value equal to 0 and a standard deviation equal to 1, via the relevant parametrization; thus, the resulting significance of IO can be conveniently estimated using widely applied thresholds ( , etc.).…”