Linking Non-Coding Variants to Genes via Chromatin Interactions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods are inadequate for identifying therapeutic targets by discovering interactions between non-coding variants and candidate genes, particularly in understanding how non-coding variants affect gene expression through long-range chromatin interactions.
Innovation Solution
A method involving the analysis of sequencing or genotyping data, chromatin accessibility data, chromatin contact profiles, and protein-chromatin binding site pairing data to identify non-coding variants located in regulatory elements, linking them to candidate genes through enhancer-promoter loops, and determining the impact on gene expression, which can be used to select therapeutics modulating signaling pathways.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current methods are used to identify therapeutic targets, then the process is simpler, but the ability to discover interactions between non-coding variants and candidate genes through long-range chromatin interactions is insufficient
Solution Approach 1:
The method segments the identification process into distinct analytical layers: (1) obtaining sequencing/genotyping data for non-coding variants, (2) obtaining chromatin accessibility data to identify regulatory elements, (3) obtaining chromatin contact profiles to detect long-range interactions, and (4) integrating these data types to link variants to candidate genes. This segmentation allows each layer to be analyzed with appropriate tools while building toward comprehensive therapeutic target identification.
Solution Approach 2:
The invention adds a spatial dimension to variant analysis by incorporating chromatin contact profile data that reveals three-dimensional chromatin architecture. This allows detection of long-range enhancer-promoter interactions that span large genomic distances, transforming the analysis from linear sequence-based methods to spatially-aware multi-dimensional analysis that captures regulatory relationships invisible to conventional approaches.
2Reliability
If multiple data types are integrated to identify long-range chromatin interactions, then the accuracy of linking non-coding variants to candidate genes improves, but the computational complexity and data processing requirements increase
Solution Approach 1:
The method performs preliminary filtering and identification steps before integration: (1) chromatin accessibility peaks are identified beforehand to define candidate regulatory elements, (2) chromatin contact profiles are pre-processed to identify significant long-range interaction pairs, and (3) non-coding variants are annotated with their genomic context. These preliminary actions reduce the search space and simplify the subsequent integration step, making the complex multi-data-type analysis more tractable.
Solution Approach 2:
Chromatin contact profile data serves as an intermediary that bridges non-coding variants and candidate genes through long-range interactions. Rather than directly comparing variant sequences to gene sequences, the method uses chromatin contact information as a mediator to identify which variants physically interact with which gene promoters, providing a mechanistic link that increases confidence in variant-gene assignments while structuring the computational analysis in manageable steps.
Data Source
AI summary
Disclosed herein are methods for linking non-coding variants and candidate genes that are associated with one another through long-range chromatin interactions. Thus, these non-coding variants may be involved in the expression of a candidate gene and therefore, serve as possible therapeutic targets for treating diseases in which the expression fo the candidate gene is dysreguated. A non-coding variant can be linked to a candidate gene by analyzing datasets including chromatin accessibility data (e.g., ATAC-seq data), protein-chromatin binding site pairing data, and/or chromatin contact profile. For example, a non-coding variant can be linked to a candidate gene by identifying an enhancer-promoter loop through a long range chromatin interaction. As another example, a non-coding variant can be linked to a candidate gene through a statistical eQTL analysis. Altogether, such links between non-coding variants and candidate genes can be used to identify novel disease genes and relevant therapies for treating related diseases.