Pattern recognition of topologically associating domains using deep learning

Abstract Background Recent increasing evidence indicates that three-dimensional chromosome structure plays an important role in genomic function. Topologically associating domains (TADs) are self-interacting regions that have been shown to be a chromosomal structural unit. During evolution, these ar...

Full description

Bibliographic Details
Main Authors: Jhen Yuan Yang, Jia-Ming Chang
Format: Article
Language:English
Published: BMC 2022-12-01
Series:BMC Bioinformatics
Subjects:
Online Access:https://doi.org/10.1186/s12859-022-05075-1
_version_ 1811179038980440064
author Jhen Yuan Yang
Jia-Ming Chang
author_facet Jhen Yuan Yang
Jia-Ming Chang
author_sort Jhen Yuan Yang
collection DOAJ
description Abstract Background Recent increasing evidence indicates that three-dimensional chromosome structure plays an important role in genomic function. Topologically associating domains (TADs) are self-interacting regions that have been shown to be a chromosomal structural unit. During evolution, these are conserved based on checking synteny block cross species. Are there common TAD patterns across species or cell lines? Results To address the above question, we propose a novel task—TAD recognition—as opposed to traditional TAD identification. Specifically, we treat Hi-C maps as images, thus re-casting TAD recognition as image pattern recognition, for which we use a convolutional neural network and a residual neural network. In addition, we propose an elegant way to generate non-TAD data for binary classification. We demonstrate deep learning performance which is quite promising, AUC > 0.80, through cross-species and cell-type validation. Conclusions TADs have been shown to be conserved during evolution. Interestingly, our results confirm that the TAD recognition model is practical across species, which indicates that TADs between human and mouse show common patterns from an image classification point of view. Our approach could be a new way to identify TAD variations or patterns among Hi-C maps. For example, TADs of two Hi-C maps are conserved if the two classification models are exchangeable.
first_indexed 2024-04-11T06:27:56Z
format Article
id doaj.art-f8432215efd744449986f19f08a4dc70
institution Directory Open Access Journal
issn 1471-2105
language English
last_indexed 2024-04-11T06:27:56Z
publishDate 2022-12-01
publisher BMC
record_format Article
series BMC Bioinformatics
spelling doaj.art-f8432215efd744449986f19f08a4dc702022-12-22T04:40:18ZengBMCBMC Bioinformatics1471-21052022-12-0122S1011510.1186/s12859-022-05075-1Pattern recognition of topologically associating domains using deep learningJhen Yuan Yang0Jia-Ming Chang1Department of Computer Science, National Chengchi UniversityDepartment of Computer Science, National Chengchi UniversityAbstract Background Recent increasing evidence indicates that three-dimensional chromosome structure plays an important role in genomic function. Topologically associating domains (TADs) are self-interacting regions that have been shown to be a chromosomal structural unit. During evolution, these are conserved based on checking synteny block cross species. Are there common TAD patterns across species or cell lines? Results To address the above question, we propose a novel task—TAD recognition—as opposed to traditional TAD identification. Specifically, we treat Hi-C maps as images, thus re-casting TAD recognition as image pattern recognition, for which we use a convolutional neural network and a residual neural network. In addition, we propose an elegant way to generate non-TAD data for binary classification. We demonstrate deep learning performance which is quite promising, AUC > 0.80, through cross-species and cell-type validation. Conclusions TADs have been shown to be conserved during evolution. Interestingly, our results confirm that the TAD recognition model is practical across species, which indicates that TADs between human and mouse show common patterns from an image classification point of view. Our approach could be a new way to identify TAD variations or patterns among Hi-C maps. For example, TADs of two Hi-C maps are conserved if the two classification models are exchangeable.https://doi.org/10.1186/s12859-022-05075-1Topologically associating domainTADHi-CChromosome organizationDeep learning
spellingShingle Jhen Yuan Yang
Jia-Ming Chang
Pattern recognition of topologically associating domains using deep learning
BMC Bioinformatics
Topologically associating domain
TAD
Hi-C
Chromosome organization
Deep learning
title Pattern recognition of topologically associating domains using deep learning
title_full Pattern recognition of topologically associating domains using deep learning
title_fullStr Pattern recognition of topologically associating domains using deep learning
title_full_unstemmed Pattern recognition of topologically associating domains using deep learning
title_short Pattern recognition of topologically associating domains using deep learning
title_sort pattern recognition of topologically associating domains using deep learning
topic Topologically associating domain
TAD
Hi-C
Chromosome organization
Deep learning
url https://doi.org/10.1186/s12859-022-05075-1
work_keys_str_mv AT jhenyuanyang patternrecognitionoftopologicallyassociatingdomainsusingdeeplearning
AT jiamingchang patternrecognitionoftopologicallyassociatingdomainsusingdeeplearning