Network and Molecular Function Analysis of Top
Weighted Liable Genes of Obstructive Sleep Apnea
Disorder Using Bioinformatics Tool
Md. Rakibul Islam1, Md. Liton Ahmed1, Bikash Kumar Paul1, 2, 3, *, Touhid Bhuiyan1,
and Kawsar Ahmed2,3
1 Department of Software Engineering, Daffodil International University
(DIU), Ashulia, Savar, Dhaka-1342, Bangladesh;
rakibul35-116@[Link]; bikash12019@[Link]
2 Department of Information and communication technology, Mawlana Bhashani Science and
Technology University, Santosh, Tangail, Bangladesh
3 Group of Bio-photomatiχ, Mawlana Bhashani Science and Technology University, Santosh, Tangail
1902, Bangladesh
* Correspondence: bikash12019@[Link], [Link]@[Link], [Link]@[Link]
Abstract: System biology network plays a crucial role to understand human diseases. Modern
Bioinformatics opens a new era of studying animals diseases. Protein-protein interaction networks
(PPINs) is one of the emergent factors for improving the cognition of biological systems. Over and above,
many physiological activities inside human bodies are demonstrated by these interaction networks.
Obstructive sleep apnea (OSA) is a serious sleep disorder that arises when a person's breathing is
disrupted during sleep. In this investigation, a computational study has been done to understand the
drug-protein interaction network and the PPI network of OSA disorder. Although this computation
method based on total 22 responsible genes ESR1, ACE, IL10, HIF1A, APP, HLADRB1, TNF, APOE, VEGFA,
IL6, MMP9, ADIPOQ, AR, CRP, PTGS2, IL1B, NOS3, CXCL8, PPARG, SLC6A4, IGF1 and LEP are noted.
Besides, molecular function from this investigation has been also demonstrated and analyzed. So, it can
be expected that this work will give an eminent contribution to the arena of the biological and
biomedical sector.
Keyword: Obstructive sleep apnea, computational biology, bioinformatics, cytoscape, protein-protein
interaction, regulatory network, drug-protein interaction.
Introduction: Biotechnology industry enhancement in last few years is novel, and development in the
pharmaceutical invention, clinical healthcare, molecular modeling, disease characterization, forensics,
and agriculture fundamentally affects economic and social issues around the world. Bioinformatics is
using the computing strategy to evaluate, handle and store biological data. Biotechnology study, in
particular incorporating sequence information handling and drug layout, happened at a rapid pace
owing to bioinformatics growth. In order to create a single field, bioinformatics combines biology,
computer science and information technology. It includes many fields of biological science, particularly
with regard to contemporary biology genomics, proteomics, genetics, transcriptomics and development.
Bioinformatics is an intriguing topic with both engineering and science inputs. Bioinformaticians are
mainly involved in the development of fresh algorithms, software, and improved databases that assist
solve many biological issues [1]. A variety of bioinformatics apps, open source or paid software and
databases are accessible to better understand biological nature and evaluate and store biological data.
Bioinformatics study is thus used to prevent the exercise of moment, price and laboratory practice [2].
Obstructive sleep apnea (OSA) symptom is a clinical disease that is often associated with excessive
snoring due to lengthy breathing phases during sleep. These interruptions split your body's oxygen
stocks for a few seconds and pull up the divergence of carbon dioxide. According to the World Health
Organization (WHO), OSA can occur many occasions during the night, making proper sleep difficult.
During the day, the person may encounter unnecessary daytime sleepiness, the trouble of
concentration, or headaches. OSA is a prevalent sleep disorder; an apnea incident is identified as the
naso-buccal airflow reprieve for more than 10 seconds [3]. Undiagnosed OSA is most widespread in
adults, the range of cruelty is extensive, and the ocular cardiovascular and behavioral symptoms in
people with OSA are also associated with undiagnosed OSA [4-6]. OSA with daytime impairment, i.e.
OSA syndrome, is projected to happen in 1 out of 20 teenagers, is normally unrecognized and untreated,
leading in the morbidity of behavior. It is reported that minimally introductory or asymptomatic OSA
results in one of five adolescents, is scarcely acknowledged and is probable to result in a high morbidity
toll at the population level [7.8]. Several analyses from the United States and Europe, including kids of
different ages from infancy through adolescence, produced minimum OSA incidence estimates of 1%-
10% [8-13].
There were so many studies on OSA linked to diagnosis, management, threat factor and medication
layout, but we collaborated on liable genes linked to OSA for the first attempt in this inquiry. In this
research, we dig out the OSA syndrome's susceptible genes. After that, we screened the top weighted
genes and by using them we generate several networks such as protein-protein association network
(PPAN), protein-protein interaction (PPI), drug-protein interaction (DPI) that will assist develop more
efficient OSA drugs in the future.
2. Methodology: Bioinformatics is often described as the implementation of computing methods to
understand and organize data related to biological components [14]. This is distinctive research on OSA
syndrome; this research discusses the relationship between proteins and will also demonstrate the
network between protein relationships with their drugs. And this research shows the PPI network's
molecular function. In this study, few procedures are conducted to reach the desired goal. These
procedures are aimed at gene processing, gene filtering, gene mining, and prevalent gene finding, etc. to
ensure the intended goal. Every single process of this study has been proved shortly below. All of these
procedures of this methodology are sub-sections inside.
2.1 Gene Collection and Filtering: Many genome databases are available for providing the data of the
genes of human, but we choose the National Center for Biotechnology Information (NCBI). NCBI has a
massive resource of biological data, information and tools. NCBI provides all the resources with free of
cost, that makes easy for research and other biological analysis. NCBI has owned 37 different massive
databases to provide various biological data [15, 16]. We use R language package "rentrez" to retrieve
the responsible genes. The "rentrez" package is used to retrieve genes or proteins from NCBI databases
of any disease [17, 18]. In the stage of gene collection, we use the "rentrez" package to collect all of
OSA's responsible genes for all of our planet's animals, but in this research, we are only working on
human OSA. So we filter the list of genes after receiving all the animal genes from our chosen disease
and keep only the human genes. This stage is very crucial for the research as it can trigger a significant
incorrect yield due to any kind of error.
2.2 Data mining and top weighted gene selection: The recorded genes from NCBI's GENE database for
the chosen disease stored in text files comprise other insignificant text to the present studies. These
insignificant documents are the trash of the research. An R language package labeled "tm" inhabits to
clear the meaningless document to prevent the issue [19]. Only the name of the genes and their
entrez ids are available to read after using the "tm" package [20]. Only top weighted and most
responsible genes are chosen for the primary study portion after this phase. Top-weighted and most
responsible genes are suitable for drug design of a disease so that only the top weighted and most liable
genes are keeps for the main analysis part.
2.3 Association and Interaction Network: Proteins are organic macromolecules capable of acting in
conjunction with several other molecules, including other proteins. They are generally acknowledged as
catalysts for metabolic pathway responses, signaling molecules in the intracellular regulation cascade,
defending proteins such as antibodies, nutrient or storage proteins, motile or contractile proteins by
their personal behavior [21, 22]. Biological interactions preserve the correct functioning of the body and
happen at all moments to guarantee the manufacturing or non-production of biomolecules engaged in
metabolism. Proteins are important components in interaction mechanisms and are responsible for
modulating catabolism and anabolism procedures with the synthesis or degradation of vital molecules,
respectively [23]. Protein-protein association network (PPAN) and protein-protein interaction network
(PPIN) both networks have represented the relationships between internal proteins. PPANs and PPINs,
also recognized as interactomes, are significant instruments for an organism's systemic research as they
constitute a collection of protein relationships whose features are prone to analyzes inherited from
graph theory. In PPAN and PPIN, it is feasible to study relationships between proteins in a complex or in
a metabolic pathway up to a number of processes, involving both low-complexity organisms as those
with elevated complexity [24]. Many online and offline tools and apps are supplied with PPINs and
PPANs which are essential for effective drug discovery. In this study, Cytoscape [25] and String [26] tools
are used to extract PPIN and PPAN. Cytoscape is an open source bioinformatics application and String is
an online database for functional biological function [24, 26].
2.4 Drug-protein Interaction Network: Protein interactions with distinct ligands generate a foundation
for an interlocked collection of vibrant procedures that provide a path of communication and regulation
within and between distinct constructions of a living organism [27]. The interaction of drugs with protein
is the primary stage of drug design. This interaction implies what types of drug are helpful for the
disease. It's the most significant output of the study. Here in this study, we use Cytoscape tool and
NetworkAnalyst [28] web interface source to design the drug-protein interaction network (DPIN).
NetworkAnalyst is an extensive web-based instrument intended to enable bench scientists to conduct
multiple prevalent and complicated gene expression meta-analysis information via an easy web
interface [28]. A minimum network is also designed from the DPN by using BisoGenet [29] plugin of
Cytoscape tool.
Result and Analysis: OSA, a common disorder projected to impact 17% of U.S. adolescents, is correlated
with enhanced danger of cardiovascular disease and general mortality. Although OSA treatment with
continuous positive airway pressure therapy is correlated with reduced general cardiovascular risk, its
effectiveness in stopping new-onset hypertension is uncertain [30-33]. The main purpose of this
research is to design the PPIN and DPN for OSA using Bioinformatics tools. Various successive
procedures are finished as mentioned in the suggested methodology to achieve the required target. The
outcomes of each method will be shown and noted in this study stage.
3.1 Gene Collection, Filtering and Selection: In the first step of this method is to retrieve all the liable
genes of OSA without applying any kind of filtering or any other preprocessing by using "rentrez"
package. The initial result of gene collection is a total of 132 genes. 132 genes belong to all the animal
species on this planet that have to endure from OSA disorder. After collecting all the genes of OSA
disorder, a filtering process applied to keep only human body genes. After the filtering process, 126
genes found which belong to the human organism. In the next step, we select the top responsible and
top weighted genes for the main analysis part. In the selection round only 22 genes are finally selected
for the rest of the study part. The final selected 22 liable and top weighted genes are ESR1, ACE, IL10,
HIF1A, APP, HLADRB1, TNF, APOE, VEGFA, IL6, MMP9, ADIPOQ, AR, CRP, PTGS2, IL1B, NOS3, CXCL8,
PPARG, SLC6A4, IGF1 and LEP.
3.2 Association and Co-expression Network: PPAN is a network that represents what kind of relation
exists between multiple proteins. This network explores the connection between multiple genes and by
this network; we can find some related results about the network. For this study, using the String
database [26] we build a PPAN for the selected 22 genes. PPAN of liable 22genes of OSA disorder
demonstrate in Fig. 1, this PPAN have a total of 177 edges. The network owns an average of 16.9 node
degree. The average local clustering coefficient of this network is 0.919, PPAN enrichment p-value of this
network is <1.0e-16. Expected number of the edge is 45 for the selected liable 22 genes. Co-expression
networks (CN) are transcript-transcript association networks, usually described as undirected graphs,
where genes are attached when there is an appreciable association of co-expression between them. CNs
are constructed from information on gene expression by calculating co-expression scores in terms of the
pairwise gene resemblance rating and selecting a meaning limit [34]. An overall CN produced by
GeneMANIA [35] prediction server in this inquiry demonstrates with Fig. 2. GeneMANIA is an open
source biological study prediction database [35, 36].
Figure 1. PPAN by STRING database. In this network 177 edges represent the association of
proteins of OSA disorder. Yellow colored edges demonstrate about text mining; pink line shows
the experimentally determined; blue line edges shows that the interaction collected from
curated database [26]. And the black line edges show the co-expression of proteins.
Figure 2. Gene co-expression network of 22 liable genes of OSA disorder by GeneMANIA [35]
prediction server. Co-expression network is an undirected graph, nodes are represents genes.
Two genes are connected with an edge if there is any momentous co-expression relationship is
found.
3.3 Protein-protein Interaction Network: PPIN is a set of protein-protein interactions that are often
stored in web databases. PPINs may complement other datasets, such as protein structural data, which
may lead to an awareness of the distinct sub-parts that add to the operation of an entire biological
system [37]. For all biological procedures, PPIN is the epitome. PPIN is usually used to show protein
interaction, and it also shows the prevalent pathway among the genes mentioned [38]. In Fig. 3 shows
the PPIN of OSA disorder. This network has a total 424 nodes and 557 edges. Network analysis of Fig. 3
demonstrates that the clustering coefficient is 0.092, network density 0.006. Network centralization of
the network is 0.312 and the characteristic path is 3.490. Fig. 4, 5 show the closeness centrality and
shared neighbors of PPI networks of OSA disorder. In Fig. 4 closeness centrality is significant between 1-
10 on the other hand, shared neighbors exists on 1 st, 2nd, 3rd, 4th, 5th, 6th and 20th positions.
Figure 3. Cytoscape generated generic protein-protein interaction network. This network owns 424
nodes and 557 edges. 424 nodes indicate individual genes, and the total amount of 557 edges exhibit
the protein-protein relationship between genes. Network density of this PPIN is 0.006.
Figure 4. Closeness centrality of PPIN of OSA Figure 5. Shared neighbors plot analysis of PPIN of
disorder. OSA disorder.
3.4 Molecular function analysis: Activities at the molecular level conducted by gene products. Terms of
molecular activity define molecular-level operations such as "catalysis" or "transport." GO molecular
function terms represent activities rather than the entities performing the actions and do not specify
where, when, or in what context the action takes place. Molecular tasks usually correspond to
operations that can be conducted by individual gene products, but some operations are conducted by
molecular complexes made up of various gene products [39]. Table 1, demonstrate the molecular
function for the responsible 22 genes. The lowest false rate for signaling receptor binding (GO:0005102)
is 1.81e-13 and highest false rate for oxidoreductase activity, acting on paired donors, with incorporation
or reduction of molecular oxygen (GO:0016705) is 0.0452.
Table 1. Molecular function analysis based of 22 liable genes PPI networks.
Molecular Function
GO-term Description Count in gene false discovery rate
set
GO:0005102 signaling receptor binding 17 of 1513 1.81e-13
GO:0048018 receptor ligand activity 10 of 458 2.28e-09
GO:0005515 protein binding 21 of 6605 9.47e-09
GO:0005125 cytokine activity 7 of 216 1.36e-07
GO:0070851 growth factor receptor binding 5 of 131 1.18e-05
GO:0005126 cytokine receptor binding 6 of 272 1.30e-05
GO:0042802 identical protein binding 11 of 1754 1.53e-05
GO:0050750 low-density lipoprotein particle 3 of 22 7.15e-05
receptor binding
GO:0046983 protein dimerization activity 9 of 1301 9.91e-05
GO:0004879 nuclear receptor activity 3 of 50 0.00055
GO:0051427 hormone receptor binding 4 of 174 0.00061
GO:0003707 steroid hormone receptor activity 3 of 59 0.00064
GO:0044877 protein-containing complex binding 7 of 968 0.00073
GO:0046914 transition metal ion binding 7 of 1051 0.0012
GO:0005496 steroid binding 3 of 88 0.0017
GO:0042803 protein homodimerization activity 6 of 830 0.0025
GO:0019899 enzyme binding 9 of 2197 0.0031
GO:0005179 hormone activity 3 of 123 0.0038
GO:0005178 integrin binding 3 of 122 0.0038
GO:0005158 insulin receptor binding 2 of 23 0.0038
GO:0071813 lipoprotein particle binding 2 of 25 0.0042
GO:0035257 nuclear hormone receptor binding 3 of 149 0.0056
GO:0042056 chemoattractant activity 2 of 33 0.0063
GO:0008201 heparin binding 3 of 161 0.0066
GO:0019904 protein domain specific binding 5 of 706 0.0072
GO:0030331 estrogen receptor binding 2 of 42 0.0091
GO:0070888 E-box binding 2 of 43 0.0092
GO:0031406 carboxylic acid binding 3 of 187 0.0092
GO:0000978 RNA polymerase II proximal 4 of 457 0.0103
promoter sequence-specific DNA
binding
GO:0008270 zinc ion binding 5 of 800 0.0104
GO:0046872 metal ion binding 11 of 4087 0.0106
GO:0044212 transcription regulatory region DNA 5 of 829 0.0113
binding
GO:0046982 protein heterodimerization activity 4 of 519 0.0132
GO:0008092 cytoskeletal protein binding 5 of 882 0.0132
GO:0070405 ammonium ion binding 2 of 66 0.0145
GO:0042277 peptide binding 3 of 270 0.0172
GO:0001664 G protein-coupled receptor binding 3 of 272 0.0172
GO:0051117 ATPase binding 2 of 80 0.0189
GO:0043167 ion binding 13 of 6066 0.0189
GO:0016209 antioxidant activity 2 of 79 0.0189
GO:0008013 beta-catenin binding 2 of 82 0.0192
GO:0008134 transcription factor binding 4 of 610 0.0195
GO:0042562 hormone binding 2 of 90 0.0221
GO:0008289 lipid binding 4 of 673 0.0246
GO:0001085 RNA polymerase II transcription 2 of 121 0.0337
factor binding
GO:0020037 heme binding 2 of 128 0.0364
GO:0001228 DNA-binding transcription activator 3 of 408 0.0391
activity, RNA polymerase II-specific
GO:0003779 actin binding 3 of 413 0.0393
GO:0016705 oxidoreductase activity, acting on 2 of 150 0.0452
paired donors, with incorporation or
reduction of molecular oxygen
Table 1. Molecular function analysis based of 22 liable genes PPI networks.
3.5 Drug-protein Interaction Network: Predicting drug-protein interactions from heterogeneous
biological data sources is an important phase in drug discovery. The challenge of this forecast
assignment resides in the rarity to be expected of recognized drug-protein interactions and countless
unidentified relationships [40]. Fig. 6 demonstrates the DPIN for OSA disorder. From the network
parameter, this network owns 400 different kinds of drugs. Network centralization is 0.174,
heterogeneity of the network 3.182 and the network density is 0.005. In Fig. 7, it’s a minimum network
of DPIN generated by BisoGenet [29] plugin of Cytoscape.
Figure 6. Drug-protein Interaction for the 22 OSA disorder genes involved. This network holds a total of
400 titles of drugs and these drugs have contact between their own and 419 edges constitute their own
relationships. This network's average diameter is 0.005.
Figure 7. Minimum network of DPIN generated by BisoGenet.
Conclusion: The human genome project and sequencing initiatives in other species have produced
unprecedented riches of biological data. The changing science of bioinformatics manages the enormous
requirement for evaluation and interpretation of this information. Bioinformatics is described as the use
of computing and analytical instruments to collect and interpret biological data [41, 42]. In this study,
we have considered genes for our selected OSA disorder from the NCBI Gene database. After some
processing, we just select 22 genes based on top weight and mostly responsible for the OSA. These 22
liable genes lead us to the ultimate analysis part. Using the genes and Bioinformatics tools we generate
PPI network, PPAN, DPI network and molecular function. These findings will help for the future analysis
for drug target networks. The upcoming investigation of this research is to design a drug target network
for the OSA disorder.
Abbreviation:
PPA = Protein-protein Association
PPAN = Protein-protein Association Network
PPI = Protein-protein Interaction
PPIN = Protein-protein Interaction Network
DPI = Drug-protein- Interaction
DPIN = Drug-protein- Interaction Network
Conflict of interest statement:
We declare that we have no conflict of interest.
Reference:
1. Goodman, N., 2002. Biological data becomes computer literate: new advances in
bioinformatics. Current opinion in biotechnology, 13(1), pp.68-71.
2. Kumar, A. and Chordia, N., 2017. Role of bioinformatics in biotechnology. Res Rev Biosci, 12(1),
pp.1-6.
3. Guilleminault C, Tilkian A, Dement WC. The sleep apnea syndromes. Annu Rev Med. 1976; 27:
465-484.
4. Strohl, K.P. and Redline, S., 1996. Recognition of obstructive sleep apnea. American journal of
respiratory and critical care medicine, 154(2), pp.279-289.
5. Pack, A.I. and Maislin, G., 2001. Who should get treated for sleep apnea?. Annals of internal
medicine, 134(11), pp.1065-1067.
6. Baumel, M.J., Maislin, G. and Pack, A.I., 1997. Population and occupational screening for
obstructive sleep apnea: are we there yet?. American journal of respiratory and critical care
medicine, 155(1), pp.9-14.
7. Wright, J. and Sheldon, T., 1998. Sleep apnoea and its impact on public health. Thorax, 53(5),
pp.410-413.
8. Young, T., Peppard, P.E. and Gottlieb, D.J., 2002. Epidemiology of obstructive sleep apnea: a
population health perspective. American journal of respiratory and critical care medicine, 165(9),
pp.1217-1239.
9. Ali, N.J., Pitson, D.J. and Stradling, J.R., 1993. Snoring, sleep disturbance, and behaviour in 4-5
year olds. Archives of disease in childhood, 68(3), pp.360-366.
10. Castronovo, V., Zucconi, M., Nosetti, L., Ossola, E., Moscatelli, D., Calori, G., Nespoli, L. and
Ferini-Strambi, L., 1998. Habitual snoring among children aged 3 to 6 years: an epidemiologic
study with objective measurement of snoring and oxygen saturation. Sleep, 21, p.S51.
11. Gislason, T. and Benediktsdottir, B., 1995. Snoring, apneic episodes, and nocturnal hypoxemia
among children 6 months to 6 years old: an epidemiologic study of lower limit of
prevalence. Chest, 107(4), pp.963-966.
12. Rosen, C.L., Arnold, J.L., DiFiore, J.M., Surovec, S.A. and Redline, S., 2000. Prevalence of key
clinical features of sleep-disordered breathing in a community-based sample of 8-10 years old US
children. Sleep, 23, p.A194.
13. Spanier-Mingoletti, S.R., Weese-Mayer, D.E. and Silvestri, J.M., 1998. Hypopnea in pediatric
polysomnography: ill-defined and underused. Am J Respir Crit Care Med, 157, p.A123.
14. Luscombe, N.M., Greenbaum, D. and Gerstein, M., 2001. What is bioinformatics? A proposed
definition and overview of the field. Methods of information in medicine, 40(04), pp.346-358.
15. Coordinators, N.R., 2017. Database resources of the national center for biotechnology
information. Nucleic acids research, 45(Database issue), p.D12.
16. Maglott, D., Ostell, J., Pruitt, K.D. and Tatusova, T., 2005. Entrez Gene: gene-centered
information at NCBI. Nucleic acids research, 33(suppl_1), pp.D54-D58.
17. Winter, D.J., 2017. rentrez: An R package for the NCBI eUtils API (No. e3179v1). PeerJ
Preprints.
18. Gentleman, R., 2008. R programming for bioinformatics. Chapman and Hall/CRC.
19. Feinerer, I., 2013. Introduction to the tm Package Text Mining in R. Accessible en ligne:
[Link] r-project. org/web/packages/tm/vignettes/tm. pdf.
20. Maglott, D., Ostell, J., Pruitt, K.D. and Tatusova, T., 2005. Entrez Gene: gene-centered
information at NCBI. Nucleic acids research, 33(suppl_1), pp.D54-D58.
21. Moraes, C.S., JUNIOR, F., Masson, G., REBELLO, K., SANTOS, L., BASTOS, N. and FARIA, R.,
2013. Série em biologia celular e molecular: Métodos experimentais no estudo de
proteínas. Fundação Oswaldo Cruz–Fiocruz, Rio de Janeiro.
22. Phizicky, E.M. and Fields, S., 1995. Protein-protein interactions: methods for detection and
analysis. Microbiol. Mol. Biol. Rev., 59(1), pp.94-123.
23. Hu, Y., Yang, Y., Fang, Z., Hu, Y.S., Zhang, L. and Wang, J., 2017. Detecting pathway
relationship in the context of human protein-protein interaction network and its application to
Parkinson’s disease. Methods, 131, pp.93-103.
24. Szklarczyk, D., Franceschini, A., Kuhn, M., Simonovic, M., Roth, A., Minguez, P., Doerks, T.,
Stark, M., Muller, J., Bork, P. and Jensen, L.J., 2010. The STRING database in 2011: functional
interaction networks of proteins, globally integrated and scored. Nucleic acids
research, 39(suppl_1), pp.D561-D568.
25. Shannon, P., Markiel, A., Ozier, O., Baliga, N.S., Wang, J.T., Ramage, D., Amin, N.,
Schwikowski, B. and Ideker, T., 2003. Cytoscape: a software environment for integrated models
of biomolecular interaction networks. Genome research, 13(11), pp.2498-2504.
26. Szklarczyk, D., Franceschini, A., Wyder, S., Forslund, K., Heller, D., Huerta-Cepas, J.,
Simonovic, M., Roth, A., Santos, A., Tsafou, K.P. and Kuhn, M., 2014. STRING v10: protein–
protein interaction networks, integrated over the tree of life. Nucleic acids research, 43(D1),
pp.D447-D452.
27. Oravcova, J., Bo, B. and Lindner, W., 1996. Drug-protein binding studies new trends in analytical
and experimental methodology. Journal of Chromatography B: Biomedical Sciences and
Applications, 677(1), pp.1-28.
28. Xia, J., Gill, E.E. and Hancock, R.E., 2015. NetworkAnalyst for statistical, visual and network-
based meta-analysis of gene expression data. Nature protocols, 10(6), p.823.
29. Martin, A., Ochagavia, M.E., Rabasa, L.C., Miranda, J., Fernandez-de-Cossio, J. and Bringas, R.,
2010. BisoGenet: a new tool for gene network building, visualization and analysis. BMC
bioinformatics, 11(1), p.91.
30. Young, T., Palta, M., Dempsey, J., Skatrud, J., Weber, S. and Badr, S., 1993. The occurrence of
sleep-disordered breathing among middle-aged adults. New England Journal of
Medicine, 328(17), pp.1230-1235.
31. Marin, J.M., Carrizo, S.J., Vicente, E. and Agusti, A.G., 2005. Long-term cardiovascular outcomes
in men with obstructive sleep apnoea-hypopnoea with or without treatment with continuous
positive airway pressure: an observational study. The Lancet, 365(9464), pp.1046-1053.
32. Young, T., Finn, L., Peppard, P.E., Szklo-Coxe, M., Austin, D., Nieto, F.J., Stubbs, R. and Hla,
K.M., 2008. Sleep disordered breathing and mortality: eighteen-year follow-up of the Wisconsin
sleep cohort. Sleep, 31(8), pp.1071-1078.
33. Punjabi, N.M., Caffo, B.S., Goodwin, J.L., Gottlieb, D.J., Newman, A.B., O'Connor, G.T.,
Rapoport, D.M., Redline, S., Resnick, H.E., Robbins, J.A. and Shahar, E., 2009. Sleep-
disordered breathing and mortality: a prospective cohort study. PLoS medicine, 6(8), p.e1000132.
34. Tieri, P., Farina, L., Petti, M., Astolfi, L., Paci, P. and Castiglione, F., 2016. Network Inference and
Reconstruction in Bioinformatics.
35. Zuberi, K., Franz, M., Rodriguez, H., Montojo, J., Lopes, C.T., Bader, G.D. and Morris, Q., 2013.
GeneMANIA prediction server 2013 update. Nucleic acids research, 41(W1), pp.W115-W122.
36. Warde-Farley, D., Donaldson, S.L., Comes, O., Zuberi, K., Badrawi, R., Chao, P., Franz, M.,
Grouios, C., Kazi, F., Lopes, C.T. and Maitland, A., 2010. The GeneMANIA prediction server:
biological network integration for gene prioritization and predicting gene function. Nucleic acids
research, 38(suppl_2), pp.W214-W220.
37. Bapat, S.A., Krishnan, A., Ghanate, A.D., Kusumbe, A.P. and Kalra, R.S., 2010. Gene
expression: protein interaction systems network modeling identifies transformation-associated
molecules and pathways in ovarian cancer. Cancer research, 70(12), pp.4809-4819.
38. Xia, J., Gill, E.E. and Hancock, R.E., 2015. NetworkAnalyst for statistical, visual and
network-based meta-analysis of gene expression data. Nature protocols, 10(6), p.823.
39. Bindea, G., Mlecnik, B., Hackl, H., Charoentong, P., Tosolini, M., Kirilovsky, A., Fridman, W.H.,
Pagès, F., Trajanoski, Z. and Galon, J., 2009. ClueGO: a Cytoscape plug-in to decipher
functionally grouped gene ontology and pathway annotation networks. Bioinformatics, 25(8),
pp.1091-1093.
40. Xia, Z., Wu, L.Y., Zhou, X. and Wong, S.T., 2010, September. Semi-supervised drug-protein
interaction prediction from heterogeneous biological spaces. In BMC systems biology(Vol. 4, No.
2, p. S6). BioMed Central.
41. Bayat, A., 2002. Bioinformatics. Bmj, 324(7344), pp.1018-1022.
42. Islam M.R., Ahmed M.L., Kumar Paul B., Asaduzzaman S., Ahmed K. (2019) Common Gene
Regulatory Network for Anxiety Disorder Using Cytoscape: Detection and Analysis. In: Rojas I.,
Valenzuela O., Rojas F., Ortuño F. (eds) Bioinformatics and Biomedical Engineering. IWBBIO
2019. Lecture Notes in Computer Science, vol 11466. Springer, Cham