coiTAD: Detection of Topologically Associating Domains Based on Clustering of Circular Influence Features from Hi-C Data

Background/Objectives: Topologically associating domains (TADs) are key structural units of the genome, playing a crucial role in gene regulation. TAD boundaries are enriched with specific biological markers and have been linked to genetic diseases, making consistent TAD detection essential. However...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Genes Jg. 15; H. 10; S. 1293
Hauptverfasser: Houchens, Drew, Chowdhury, H. M. A. Mohit, Oluwadare, Oluwatosin
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Switzerland MDPI AG 01.10.2024
MDPI
Schlagworte:
Algorithms
Biomarkers
Chromatin
Chromatin - genetics
Chromatin Assembly and Disassembly
Cluster Analysis
Clustering
Comparative analysis
Computational Biology - methods
Gene regulation
Genetic aspects
Genetic disorders
Genomes
Genomic analysis
Humans
Identification
Identification and classification
Nucleotide sequence
Physiological aspects
Proteins
Structure
United States
features
clustering
topologically associating domain
TAD caller
principal component analysis
ISSN:2073-4425, 2073-4425
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Background/Objectives: Topologically associating domains (TADs) are key structural units of the genome, playing a crucial role in gene regulation. TAD boundaries are enriched with specific biological markers and have been linked to genetic diseases, making consistent TAD detection essential. However, accurately identifying TADs remains challenging due to the lack of a definitive validation method. This study aims to develop a novel algorithm, termed coiTAD, which introduces an innovative approach for preprocessing Hi-C data to improve TAD prediction. This method employs a proposed “circle of influence” (COI) approach derived from Hi-C contact matrices. Methods: The coiTAD algorithm is based on the creation of novel features derived from the circle of influence in input contact matrices, which are subsequently clustered using the HDBSCAN clustering algorithm. The TADs are extracted from the clustered features based on intra-cluster interactions, thereby providing a more accurate method for identifying TADs. Results: Rigorous tests were conducted using both simulated and real Hi-C datasets. The algorithm’s validation included analysis of boundary proteins such as H3K4me1, RNAPII, and CTCF. coiTAD consistently matched other TAD prediction methods. Conclusions: The coiTAD algorithm represents a novel approach for detecting TADs. At its core, the circle-of-influence methodology introduces an innovative strategy for preparing Hi-C data, enabling the assessment of interaction strengths between genomic regions. This approach facilitates a nuanced analysis that effectively captures structural variations within chromatin. Ultimately, the coiTAD algorithm enhances our understanding of chromatin organization and offers a robust tool for genomic research. The source code for coiTAD is publicly available, and the URL can be found in the Data Availability Statement section.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:2073-4425
2073-4425
DOI:10.3390/genes15101293