=Paper=
{{Paper
|id=Vol-2807/shortP
|storemode=property
|title=OGG-CoV: Ontology Representation and Analysis of Genes and Genomes of Coronaviruses (short paper)
|pdfUrl=https://ceur-ws.org/Vol-2807/shortP.pdf
|volume=Vol-2807
|authors=Anthony Huffman,Yongqun He
|dblpUrl=https://dblp.org/rec/conf/icbo/HuffmanH20
}}
==OGG-CoV: Ontology Representation and Analysis of Genes and Genomes of Coronaviruses (short paper)==
https://ceur-ws.org/Vol-2807/shortP.pdf
OGG-CoV: Ontology Representation and
Analysis of Genes and Genomes of
Coronaviruses
Anthony Huffman a,1 and Yongqun He a
a
University of Michigan Medical School Ann Arbor, MI 48109, USA
Abstract. The current SARS-CoV-2 pandemic has brought about significant influx
of coronavirus data for better disease understanding and treatment and vaccine
development. As such, ontologies to help categorize the massive amount of research
and information being done are required. The Ontology of Genes and Genomes
(OGG) systematically represents genes and genomes for specific organisms. OGG-
CoV is a branch of OGG that provides an ontological representation of genes and
genomes within different coronavirus strains and species. OGG-CoV adopts a pan-
genome strategy and systematically represents coronavirus genes based on their
ortholog classification.
Keywords. Ontology, Coronavirus, Genes, Genomes
1. Introduction
Coronaviruses are a subfamily of RNA viruses, some of which are responsible for
various human respiratory illnesses [1]. The latest, SARS-COV-2, has caused the
COVID-19 pandemic that, as of 08/17/2020, has reached over 21,707,773 confirmed
cases in 188 countries and has prompted a massive expansion of genomic sequencing to
support a better understanding of the disease and rational development of effective and
safe treatments and cures (https://coronavirus.jhu.edu/us-map). The NIH already has
16,084 sequences uploaded for this virus and there are 23,295 articles on PubMed written
within the last eight months, demonstrating the need to categorize the massive data influx.
Coronavirus genes are generally divided into structural proteins and accessory
proteins. The structural proteins are responsible for the formation of the virion and its
transmission and have well-characterized functions and roles [2]. Accessory proteins
form parts of the virion and vary more often between species, with some being linked as
contributing to the coronavirus pathology [3]. RNA viruses are highly mutable, with
SARS-CoV-2 having mutation rates of ~30% within some of their genes [4].
Ontologies are a tool used to classify and systematize the information of entities and
relations to aid further research and inquiry. The Ontology of Genes and Genomes
(OGG) is an ontology developed to classify the genes and genomes of organisms, with
sub-ontologies developed for specific model species [5]. The Coronavirus Infectious
Disease Ontology (CIDO) (https://github.com/CIDO-ontology/) is a community-based
1
Anthony Huffman, Corresponding author, University of Michigan Medical School, Ann Arbor, MI, USA. E-
mail: huffmaar@umich.med.edu,
Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
ontology in the domain of coronavirus diseases, which covers various topics such as virus
etiology, hosts, host-virus interactions, drugs and vaccines [6]
In this study, we present OGG-CoV, an extension of the OGG core with a focus on
logically modeling and representing genes and genomes in various coronaviruses. OGG-
CoV has then been imported into the CIDO to support integrative representation and
analysis of coronavirus genes and their usages in different applications.
2. Methods
2.1. Coronavirus gene and genome information extraction
The genes and genetic sequences were all collected and extracted from NCBI. Sections
of polyprotein genes that produce functional gene products were assigned from UniProt’s
InterPro.
2.2. OGG-CoV development
OGG-CoV was developed by aligning it with OGG and CIDO [7]. Orthologs between
coronaviruses were determined using the NCBI gene annotation and literature annotation
for genes with known protein products. Genes that do not have labeled orthologs were
classified using nucleotide alignments from MUSCLE
(https://www.ebi.ac.uk/Tools/msa/) and ViPR BLAST (https://www.viprbrc.org/brc/-
home.spg?decorator=vipr). If both programs predicted a different set of orthologs or
there was no match, the gene was placed within its own class. The basic OGG terms and
structure were obtained using Ontofox [6]. Extracted genome and gene data were
uploaded using Ontorat into version 5.5 of Protégé-OWL editor
(https://protege.stanford.edu/) for editing.
2.3. OGG-CoV source code
The source code of OGG-CoV is available at GitHub: https://github.com/CIDO-
ontology/OGG-CoV. The source code uses the license CC-BY.
3. Results
3.1. High level ontology design
Figure 1 lays out the high level design, using Basic Formal Ontology (BFO) inherited
from the high level design of OGG. The OGG-CoV is designed to represent specific
.
Figure 1. Screenshot of high level of design of OGG-CoV. Each gene from a specific strain is a subclass for
the gene within a species and the gene within the strain. The origin of each terms is shown below each class in
parenthesis: BFO: Basic Formal Ontology, PR: Protein Ontology, OGG: Ontology of Genes and Genomes,
NCBITaxon: NCBITaxon.
genes (e.g., orf1ab gene of SARS-CoV-2 Wuhan strain) under the gene of its own strain
(e.g., gene of SARS-CoV-2 Wuhan strain) and meanwhile under its evolution-derived
ortholog hierarchy (e.g., coronavirus S gene). This design asserts S gene of SARS-CoV-
2 Wuhan strain under two parent classes. However, this design allows us to compare
different genes within an ortholog.
3.2. Specific OGG-CoV development
In our current OGG-CoV development, we included 13 strains from 8 different species.
SARS-CoV-2 Wuhan strain, the first sequenced SARS-CoV-2 strain, has 10 genes. As
seen in Figure 2, the coronavirus genes are classified based on their taxonomic
classification and orthology gene type. For example, the S gene of SARS-CoV-2 Wuhan
strain can be classified as a gene under different strains (e.g., Wuhan strain) (Figure 2a)
or as a gene under an ortholog S gene type (Figure 2b).
The orf1ab gene is made up of two different genes, orf1a and orf1b. Orf1a gene
encodes a single polypeptide which produces 11 specific nonstructural proteins (nsps)
Orf1b produces a polypeptide that which produces all but the last nsp produced by orf1a
in addition to 5 unique nsps responsible for viral replication. This is due to a frame shift
caused by ribosomal slippage during translation of orf1a. Together they code for the
coronavirus replicase polyprotein. Each nsp-encoding portion of this replicase gene is
represented in ontology as a part of the polygene sequence which encodes a specific nsp
protein.
4. OGG-CoV Use case: Creation of a Coronavirus Pan-genome
Pan-genome is defined as the entire set of genes for all strains within a clade (i.e., a group
of organisms evolved from a common ancestor) [9]. While originally restricted to
bacteria, this can be expanded to other organisms, including viruses. A pan-genome
typically includes a set of core genes that are present in all strains with the clade and
accessory genes that are present in some but not all strains. Our OGG-CoV is designed
to represent the genes in all the coronavirus clade based on the ortholog analysis, which
allows pan-genome representation.
Figure 2. Classification of coronavirus genes in OGG-CoV. Each class that contains a specific s gene is sorted
as a subclass for its gene taxonomy origin (A) and gene type (B). Orf1a and orf1b were labeled as a single gene,
orf1ab, on the NCBI website. Each gene is purposely designed to have two explicit parents. The genes circled
in red are classified as S genes due to NCBI listing a spike polypeptide product for these genes.
The selected coronaviruses genomes had a consistent pattern of the orf1ab gene, and
the S (spike), E (envelope), M (membrane), and N (nucleocapsid) structure genes
occurring in this specific order from the 5’ to 3’ direction. The one exception, AIBV
NCBI reference genome, did not have a gene labeled as an M protein but did have an
unlabeled gene between the E and N proteins. The orf1ab always demonstrated a -1
frameshift which coded for a single polypeptide that is cleaved to create 13-16 total
nonstructural proteins, depending on the species. This seems to indicate that the
coronavirus pan-genome has these structural proteins has its core genes.
The accessory genome of coronaviruses contains some structures genes that are part
of the core genome for certain coronavirus sub-clades and the various accessory genes.
We found that the structural gene HE (hemagglutinin-esterase) was found to be part of
the core genome subgenus Embecovirus but not present in other coronavirus species. The
various accessory genes differed in quantity between each species and were labeled by
their ordinal location within the genome from the 5’ to 3’ direction. This nomenclature,
while allowing for quick comparison of the number of accessory genes between
coronaviruses, is insufficient to match accessory proteins within the ontology without
supporting evidence from sequence alignment.
Interestingly, we found the name of the orf10 gene only in SARS-CoV-2 strains but
not in other branches of coronaviruses. However, a unique name does not guarantee its
unique presence among all coronavirus genes. To address this issue, we performed a
MUSCLE analysis, and found that orf10 had no significant alignment with any other
coronavirus proteins, suggesting that orf10 is uniquely present in SARS-CoV-2. We will
later include more genomes from different coronavirus strains to confirm the unique
presence of this gene in SARS-CoV-2. If the conclusion is true, orf10 likely plays an
important role in making SARS-CoV-2 unique in terms of the viral pathogenesis and/or
transmission.
5. Discussion and Conclusion
OGG-CoV is developed to serves as a knowledge base of coronavirus genes and genomes.
Instead of presenting genes and genomes in a single species as done in previous OGG
branches, OGG-CoV adopts a pan-genome strategy to systematically represent
coronavirus genes not only in terms of its genomes, but also based on their ortholog
classification. The ortholog classification method will then allow us to easily define the
core and accessory genes of the pan-genome of all coronaviruses. Our use case
demonstrates that many structure genes belong to the core genes. However, some genes
such as the orf10 gene is unique is only SARS-CoV-2 strain.
Future work will include its integration with the other parts of CIDO and allow the
other entities (e.g., vaccine and drug target) defined in CIDO to possibly link to the genes
and genomes in OGG-CoV. We will also expand OGG-CoV to possibly identify and
represent gene mutations that occur naturally in various coronaviruses, allowing for
mechanistic analysis of host-virus interactions and fight against COVID-19 pandemic.
Acknowledgments
Edison Ong and Lauren Austin for their aid and commentary. This was funded by the
ImmPort project (NIH NIAID grant 1UH2AI132931).
References
[1] Cui J, Li F, Shi ZL. Origin and evolution of pathogenic coronaviruses. Nat Rev Microbiol. 2019
Mar;17(3):181-192. doi: 10.1038/s41579-018-0118-9. PMID: 30531947; PMCID: PMC7097006.
[2] Fehr AR, Perlman S. Coronaviruses: an overview of their replication and pathogenesis. Methods Mol
Biol. 2015;1282:1-23. doi: 10.1007/978-1-4939-2438-7_1. PMID: 25720466; PMCID: PMC4369385.
[3] Yue Y, Nabar NR, Shi CS, Kamenyeva O, Xiao X, Hwang IY, Wang M, Kehrl JH. SARS-Coronavirus
Open Reading Frame-3a drives multimodal necrotic cell death. Cell Death Dis. 2018 Sep 5;9(9):904.
doi: 10.1038/s41419-018-0917-y. PMID: 30185776; PMCID: PMC6125346.
[4] Pachetti M, Marini B, Benedetti F, Giudici F, Mauro E, Storici P, Masciovecchio C, Angeletti S,
Ciccozzi M, Gallo RC, Zella D, Ippodrino R. Emerging SARS-CoV-2 mutation hot spots include a
novel RNA-dependent-RNA polymerase variant. J Transl Med. 2020 Apr 22;18(1):179. doi:
10.1186/s12967-020-02344-6. PMID: 32321524; PMCID: PMC7174922.
[5] He Y, Liu Y, Zhao B. (2014) OGG: a Biological Ontology for Representing Genes and Genomes in
Specific Organisms. In: Hogan WR, Arabandi S, Brochhausen, M, editor. Proceedings of the 5th
International conference on Biomedical Ontologies; October 8-9 2014; Houston, TX. Published on
CEUR-WS; c2014 p. 13–20
[6] He Y, Yu H, Ong E, Wang Y, Liu Y, Huffman A, Huang HH, Beverley J, Hur J, Yang X, Chen L,
Omenn GS, Athey B, Smith B. CIDO, a community-based ontology for coronavirus disease knowledge
and data integration, sharing, and analysis. Sci Data. 2020 Jun 12;7(1):181. doi: 10.1038/s41597-020-
0523-6. PMID: 32533075; PMCID: PMC7293349.
[7] Xiang Z, Courtot M, Brinkman RR, Ruttenberg A, He Y. OntoFox: web-based support for ontology
reuse. BMC Res Notes. 2010 Jun 22;3:175. doi: 10.1186/1756-0500-3-175. PMID: 20569493; PMCID:
PMC2911465.
[8] Xiang Z, Zheng J, Lin Y, He Y. Ontorat: automatic generation of new ontology terms, annotations, and
axioms based on ontology design patterns. J Biomed Semantics. 2015 Jan 9;6:4. doi: 10.1186/2041-
1480-6-4. PMID: 25785185; PMCID: PMC4362828.
[9] Computational Pan-Genomics Consortium. Computational pan-genomics: status, promises and
challenges. Brief Bioinform. 2018 Jan 1;19(1):118-135. doi: 10.1093/bib/bbw089. PMID: 27769991;
PMCID: PMC5862344.