CCIP: predicting CTCF-mediated chromatin loops with transitivity

Abstract

Motivation CTCF-mediated chromatin loops underlie the formation of topological associating domains (TADs) and serve as the structural basis for transcriptional regulation. However, the formation mechanism of these loops remains unclear, and the genome-wide mapping of these loops is costly and difficult. Motivated by the recent studies on the formation mechanism of CTCF-mediated loops, we studied the possibility of making use of transitivity-related information of interacting CTCF anchors to predict CTCF loops computationally. In this context, transitivity arises when two CTCF anchors interact with the same third anchor by the loop extrusion mechanism and bring themselves close to each other spatially to form an indirect loop. Results To determine whether transitivity is informative for predicting CTCF loops and to obtain an accurate and low-cost predicting method, we proposed a two-stage random-forest-based machine learning method, CCIP (CTCF-mediated Chromatin Interaction Prediction), to predict CTCF-mediated chromatin loops. Our two-stage learning approach makes it possible for us to train a prediction model by taking advantage of transitivity-related information as well as functional genome data and genomic data. Experimental studies showed that our method predicts CTCF-mediated loops more accurately than other methods and that transitivity, when used as a properly defined attribute, is informative for predicting CTCF loops. Furthermore, we found that transitivity explains the formation of tandem CTCF loops and facilitates enhancer-promoter interactions. Our work contributes to the understanding of the formation mechanism and function of CTCF-mediated chromatin loops. Availability and implementation The source code of CCIP can be accessed at: https://github.com/GaoLabXDU/CCIP. Supplementary information Supplementary data are available at Bioinformatics online.

Inferring CTCF binding patterns and anchored loops across human tissues and cell types

“…Although many computational models that can predict CTCF-mediated loops with varied sensitivity and specificity (Oti et al 2016; Kai et al 2018; Matthews and Waxman 2018; Zhang et al 2018; Ibn-Salem and Andrade-Navarro 2019; Cao et al 2021; Kuang and Wang 2021; Lv et al 2021; Wang et al 2021; Xi and Beer 2021), they usually learn from CTCF ChIA-PET data and barely evaluate the loop attribute among different types of CBS. Here, we applied cohesin ChIA-PET data and strict CTCF binding evidence to specifically analyze CTCF binding at the insulator and its anchored loop, which provides a unique tool for characterizing loop formation through the loop extrusion mechanism.…”

“…Although many computational models that can predict CTCF-mediated loops with varied sensitivity and specificity (Cao et al, 2021; Deng et al, 2022; Ibn-Salem and Andrade-Navarro, 2019; Kai et al, 2018; Kuang and Wang, 2021; Lv et al, 2021; Matthews and Waxman, 2018; Oti et al, 2016; Wang et al, 2021; Xi and Beer, 2021; Zhang et al, 2018), they usually learn from CTCF ChIA-PET data and barely evaluate the regulatory potential and loop attribute among different types of CBS. Here, we applied cohesin ChIA-PET/ChIP-seq data, ChromHMM-predicted insulator, and strict CTCF binding evidence to specifically analyze CTCF binding patterns at three types of CTCF-mediated CREs.…”

“…However, these studies are either unable to count the feature patterns of individual CBSs or limited to given genomic loci. Besides, a large number of computational tools were developed to qualitatively or quantitatively predict CTCF-mediated loops (Cao et al, 2021; Ibn-Salem and Andrade-Navarro, 2019; Kai et al, 2018; Kuang and Wang, 2021; Lv et al, 2021; Matthews and Waxman, 2018; Oti et al, 2016; Wang et al, 2021; Xi and Beer, 2021; Zhang et al, 2018), but few could specially evaluate the regulatory potential of DNA sequence at CBS and how it affects the loop formation. Moreover, it remains elusive to see which factors or combinations of features determine the specificity among different types of functional CTCF binding events, and whether such information could contribute the prediction of CTCF-anchored loops across the whole genome.…”

Commonly used software tools produce conflicting and overly-optimistic AUPRC values

“…This linear interpolation is used when computing area under the curve with the trapezoidal rule in auc." Despite this cautionary note, some recently published studies still used this method to compute AUPRC [39][40][41][42][43][44].…”

“…On the manual page about its model evaluation modules (https://scikit-learn.org/stable/modules/model_evaluation.html#precision-recall-f-measure-metrics), it is mentioned that the linear interpolation method can lead to overly-optimistic AUPRC values: “References [Davis2006] and [Flach2015] describe why a linear interpolation of points on the precision-recall curve provides an overly-optimistic measure of classifier performance. This linear interpolation is used when computing area under the curve with the trapezoidal rule in auc.” Despite this cautionary note, some recently published studies still used this method to compute AUPRC [39–44].…”