=Paper=
{{Paper
|id=Vol-1320/paper_24
|storemode=property
|title=GSO: A Semantic Web Ontology to Capture 3D Genome Structure
|pdfUrl=https://ceur-ws.org/Vol-1320/paper_24.pdf
|volume=Vol-1320
|dblpUrl=https://dblp.org/rec/conf/swat4ls/ZhengDTT14
}}
==GSO: A Semantic Web Ontology to Capture 3D Genome Structure==
https://ceur-ws.org/Vol-1320/paper_24.pdf
GSO: A Semantic Web Ontology to Capture 3D Ge-
nome Structure
W. Jim Zheng1, Jingcheng Du1, Jijun Tang2, Cui Tao1
1
School of Biomedical Informatics
University of Texas Health Science Center at Houston
{wenjin.j.zheng,jingcheng.du,cui.tao}@uth.tmc.edu
1
Department of Computer Science
University of South Carolina
{jtang}@cse.sc.edu
Abstract. More and more evidences indicate that the 3D conformation of eu-
karyotic genome plays important functional role in the cell. While extensive
experimental investigations have been performed to study such structure-
function relationships using techniques such as Hi-C methods, there is no data
standard to capture the 3D conformation of the genome. In previous work, we
have developed an object-oriented framework, Genome3D, to integrate, model
and visualize human genome in 3-dimension. We report here about a high level
ontology, Genome Structure Ontology (GSO), to capture the 3D conformation
of eukaryotic genome. GSO captures the structural organization of eukaryotic
genome at levels ranging from chromosome, to 30nm fiber, to nucleosome and
then to the atomic level of DNA. We believe such ontology could play a signif-
icant role in annotating, analyzing and describing molecular data for the 3D
conformation of eukaryotic genome.
1 Introduction
The latest research indicates that spatial conformation and interaction of chromatin
(1, 2) play a fundamental role in important cellular functions (such as gene regulation)
and cell state determination (e.g., stem cell pluripotency). The influx of new details
about the higher-level structure and dynamics of the genome enabled many recent
efforts measuring and modeling genome structures at various resolutions and levels of
detail (3-10). While incorporating this spatial information can yield a 3D genome
model, the size and complexity of these data dictate the design and development of
new algorithms and methods in data integration and model construction.
The 3D conformation of eukaryotic genome is complex: at high level, chromo-
somes go through different states through the cell cycle and take different confor-
mations; each of these states will have different structures; at low level, the confor-
mation from 30nm fiber to nucleosome, and to histone and DNA are also affected by
the cell cycle. Precise modeling of these chromosomes’ conformation is critical for
us to understand the structure-function relationships for the 3D conformation of the
genome. As the exploration of the genome 3D structure becomes more and more
comprehensive, developing a Genome Structure Ontology (GSO) to capture the struc-
tural information in formal specifications becomes imperative. While there are ex-
tensive ontologies, such as gene ontology and sequence ontology that have been de-
veloped to model genome information, none is aimed at capturing the 3D genome
structure.
We created the first model-view framework of eukaryotic genomes, Genome3D, to
enable integration and visualization of genomic and epigenomic data in a three-
dimensional space (11). Our model of the physical genome implicitly contains all
levels of structure and hierarchy, and provides an underlying platform for integrating
multi-scale genomic information within three dimensions. Our viewer uses a hierar-
chical model of the relative positions of all nucleotide atoms in the cell nucleus.
Through this work, we have gained extensive experience in capturing the multi-scale
information about the 3D structure of human genome. Here we report our initial
attempt to transform this knowledge into Genomic Structure Ontology at high level.
2 Methods and Results
We created the GSO in Web Ontology Language (OWL) (12) in the Protégé ontol-
ogy editing environment. OWL is built on formalisms that use Description Logic
(DL)(13) to allow reasoning and inference. The Rule Interchange Format (RIF)(14)
can be used to add rules to OWL and can be used to infer new knowledge from an
OWL based ontology and reason about OWL individuals. Moreover, the Semantic
Web community has developed open source tools for editing, storing, and reasoning
over information represented in OWL. Building the GSO in OWL will allow us to
directly OWL’s formalism and tools for semantic defining the knowledge in the do-
main and perform reasoning in the future.
Figure 1 shows the hierarchy defined in GSO. The hierarchy reflects the conceptual
structure of eukaryotic chromosome, together with high level, critical information
about the state of the chromosome and the cell cycle phases, which are critical in
determining the 3D conformation of the chromosome. The details for the anaphase
chromosome are captured with details all the way down to the DNA. The design of
this hierarchy is accommodative such that the details for other phases or other types
of genomes can be incorporate in the future.
GSO provides a way for us to semantically define important genome structure entities
such as Nucleosome Core Particles (NCP). Each NCP has a 146 bps long DNA wrap-
ping around the histone core, with a radius and rise, and has a unique position and
orientation define by a radial vector and an axis. The histone core is an octamer with
two copies of Histone H2A, H2B, H3 and H4. Figure 2 shows the ontological repre-
sentation of the NCP. Each NCP has a base pair (BP) index, which is composed of
three vectors, axis of radiation, center position, and the radius to the starting BP. We
created a new class called BPIndex and three object properties (axisofRotation, cen-
terPosition, and radiusToStartingBP) to represent the three vectors. Each vector is
Figure 1: Hierarchy of genome structure for eukaryotic genome defined in GSO.
represented by a class called 3D_Stucture. The domain of the object properties is
BPIndex and the range is 3D_Stucture. Three data properties x_coord, y_coord, and
z_coord have been defined to represent the x, y, and z coordinator values in a 3D
structure. In addition, we also created properties to represent other important features
of an NCP. The Data property rise_nm is defined to represent the rise per turn of the
NCP in nm; the Data property rot_rad is defined to represent the length of DNA spi-
ral around the NCP in radians; the Data property ncp_bps is defined to represent the
number of bps in the DNA NCP wrap; and the Data property is defined to represent
the number of NCPs in file. After these classes and properties have been defined, we
can represent individual NCPs in RDF with respective of the GSO definition. Figure 3
shows an example RDF representation of an NCP (partial).
Figure 2: Ontological definition of NCP
Figure 3: RDF representation of a sample NCP (partial). Each NCP is defined by
a set of parameters: rise, rotation, ncp bps and BPIndex. The BPIndex is defined by
center position, the axis of rotations and the radius of starting base pair
GSO is intended to capture the 3D conformation of the genome, and has overlaps
with other existing ontologies. In order to avoid re-inventing the wheel, we surveyed
the existing ontologies for ontology terms that we used in the GSO. We applied the
NCBO annotator on the all the GSP terms and searched for matches for all the ontolo-
gies hosted by the NCBO Bioportal. Not surprisingly, we have found that 98 other
ontologies contain terms in GSO. We have created a matrix to link the terms in GSO
with the terms in the other ontologies. Out of the 98 ontologies, the top 10 ontologies
are BIOMODELS, CRISP, MESH, RH-MESH, NIFSTD, AURA, GO, GO-EXT,
HINO, and SNML. We will further investigate these ontologies and terminologies
and align the GSO terms to them when applicable.
3 Discussion
Built upon our knowledge of constructing a 3D physical model of the human genome,
we developed a Genome Structure Ontology (GSO) for eukaryotic genome. GSO
captures the basic structure of the genome from the whole genome level to the chro-
mosome level, to the 30nm fibre level, to the nucleosome level and then down to the
atomic level of the DNA structure. Combined with quantitative parameters, GSO also
defines the details of genome structure at each level.
A mature GSO cannot only define a 3D genomic model, but also annotating exist-
ing data used to construct the model. For example, analysis results from Chip-Seq
data for the nucleosome position can be annotated with the ontology term such as
NCP. Such annotation can help data user to quickly identify data needs to be used for
3D model construction. While annotating a 3D model of a eukaryotic genome, the
ontology can help to infer the relationships of different components of a model, thus
providing useful insight from spatial relationships. For example, in our original Ge-
nome3D paper (11), we gave an example of multiple Single Nucleotide Polymor-
phisms (SNPs) occurred within an NCP, we can easily infer all the occurrence of such
SNPs for the whole genome using GSO. Another example of such application is an
enrichment analysis of ontology terms in space: given a transcription hotspot or
CHIA-PET (Chromatin Interaction Analysis by Paired-End Tag Sequencing) spot,
what are the ontology terms are enriched within a radius of 0.1μm.
Even though the GSO is still a prototype, further refinement of this prototype could
make it a very useful tool to annotate and to capture the 3D structure of eukaryotic
genome. More semantic annotations can be built on top of the GSO classes in order to
support automatic inference for different use cases.
4 Acknowledgement
This work is supported by NSF award #1339470 to WJZ, #1161586 to JT, and
NIH/NLM award R56 LM010680 to WJZ and R01LM011829 to JD and CT.
References:
1. Fullwood MJ, Liu MH, Pan YF, Liu J, Xu H, Mohamed YB, et al. An
oestrogen-receptor-alpha-bound human chromatin interactome. Nature.
2009;462(7269):58-64. Epub 2009/11/06.
2. Lieberman-Aiden E, van Berkum NL, Williams L, Imakaev M, Ragoczy T,
Telling A, et al. Comprehensive mapping of long-range interactions reveals folding
principles of the human genome. Science. 2009;326(5950):289-93. Epub 2009/10/10.
3. P. Hahnfeldt JEH, D.J. Brenner, R.K. Sachs, and Lynn R. Hlatky. Polymer
Models for interphase chromosomes. PNAS. 1993;90(16):7854-8.
4. Sachs RK, van den Engh G, Trask B, Yokota H, Hearst JE. A random-
walk/giant-loop model for interphase chromosomes. Proc Natl Acad Sci U S A.
1995;92(7):2710-4.
5. Ponomarev AL, Brenner D, Hlatky LR, Sachs RK. A polymer, random walk
model for the size-distribution of large DNA fragments after high linear energy
transfer radiation. Radiat Environ Biophys. 2000;39(2):111-20.
6. Woodcock CL, Grigoryev SA, Horowitz RA, Whitaker N. A chromatin
folding model that incorporates linker variability generates fibers resembling the
native structures. Proc Natl Acad Sci U S A. 1993;90(19):9021-5.
7. Dekker J, Rippe K, Dekker M, Kleckner N. Capturing chromosome
conformation. Science. 2002;295(5558):1306-11.
8. Balaeff A, Mahadevan L, Schulten K. Modeling DNA loops using the theory
of elasticity. Phys Rev E Stat Nonlin Soft Matter Phys. 2006;73(3 Pt 1):031919.
9. Beard DA, Schlick T. Computational modeling predicts the structure and
dynamics of chromatin fiber. Structure. 2001;9(2):105-14.
10. Sharma S, Ding F, Dokholyan NV. Multiscale modeling of nucleosome
dynamics. Biophysical journal. 2007;92(5):1457-70.
11. Asbury TM, Mitman M, Tang J, Zheng WJ. Genome3D: a viewer-model
framework for integrating and visualizing multi-scale epigenomic information within
a three-dimensional genome. BMC Bioinformatics. 2010;11:444. Epub 2010/09/04.
12. McShan DC, Rao S, Shah I. PathMiner: predicting metabolic pathways by
heuristic search. Bioinformatics. 2003;19(13):1692-8.
13. Horrocks I, Patel-Schneider PF, McGuinness DL, Welty CA. OWL: a
Description Logic Based Ontology Language for the Semantic Web. The Description
Logic Handbook: Theory, Implementation, and Applications 2ed: Cambridge
University Press; 2007.
14. van Helden J, Naim A, Mancuso R, Eldridge M, Wernisch L, Gilbert D, et al.
Representing and analysing molecular and cellular function using the computer. Biol
Chem. 2000;381(9-10):921-35.