International Journal of Trendy Research in Engineering and Technology

Volume 4 Issue 7 December 2020

ISSN NO 2582-0958
________________________________________________________________________________________________

NETWORK OF DISEASES AND ITS ENDOWMENT TOWARDS DISEASE

UNDERSTANDING

Tripathi Rajesh Kajal

Maharashtra institute of Pharmacy, Pune, Maharashtra
[email protected]

ABSTRACT
Disease Network is the science that has emerged to diagnose a disease from a network aspect
specifically. Networks are the group that interconnect to each others similarly disease networks are
the one that reveal concelled connection among apparently independent biomedical entities like
physiologic process, signaling receptors, in addition to genetic code, also they prove to exists
intitutive in addition to powerful way to learn/discover or diagnose a disease.Due to these networks,
we can now consume the elderly drugs and its method to learn/discover the new drug
accordingly.Example- Colchicine is used in gout but after repurposing it is also used in mediterranean
fever. This is because there are many factors that affect the body during mediterranean fever and
gout, we know that gout is a form of arthritis that causes pain in joints also mediterranean fever is the
one which is accompanied by pain in joints, therefore colchicine is used as a repurposed drug again.In
repurposing of medicines or drugs we first analyse the change in symptoms and identify the target
organ and accorgingly we produce a drug that is compatible with pharmacokinetics of the body. As
the availablity of transcriptomic,proteomic and metabolomic data sources are increasing day by day it
helps in classification of disease .Also there are some networks reffered to as complex networks
which can be called as collection of linked junctions/ nodes.

1.INTRODUCTION observed.There is a graph theory in network

We understand disease when we diagnose it theory which relates to nodes and edges having
through symptoms. And when we understand attributes. Network theory is a part of graph theory
the disease we get to know about the particular where graph is made by vertices connected by
bacteria,or virus or any antigen affecting the edges and network is made by nodes connected by
body harmfully, also we get to know about links.
various factors affecting the bacteria or virus in
a particular tempreture/pressure etc therefore Actually network perspective is to look beyond
we design an antibody which helps us to get formal, and simple designated relationships to
rid of bacteria or virus, but we cant forget that complex connections between people, (if we call
disease networks play a very important role in terms of eukaryotic beings) but when we talk
here. Disease netorks are the one that tell us about bacteri,virus or germs then the network
about the understanding of disease. Only between them and the body is very complex.
through these networks we come to know Complex network theory is an interpretation of
about the disease. As we know due to statistical data (physics) of old graph theory aimes
functional interdependencies in molecular at description and understanding the composition
compounds there arise abnormality in gene but that was created by connections between elements
this is rarest because initially the molecular of a complex system. These elements are
compounds only deals with other cellular connected by junctions/nodes in a pairwise
components except genes.and show their manner and in links whenever a relationship is
symptoms but when heriditary is considered observed between corrosponding elements. The
some of the disease come from gene where the resulting composition can be described by many
abnormality arises from phenotypic character topological metrics and can used as the base for
of previous one. reshaping the system.
However ,networks define symtoms which have Also disease networks expresses connections
important role to play in every disease. Therefore among junctions and edges in a graph, where
network defines symptoms. Scientists named G=(D,W), D represents set of disease called as
Erdos and Renyi created the network theory based junction/ nodes and W represents set of
on the mathematical model which was used to connections/ relationships ( edges) based on
scrutinize the structural properties of a random similarity.Here meaning of similarity varies
network where pair of nodes were connected and depending on data used to build the network
an obvious expectation was observed which was which can be biologicsl or phenotyphic among
reresented in the form of probability was other approaches. Therefore, through this article

www.trendytechjournals.com 35
 International Journal of Trendy Research in Engineering and Technology
Volume 4 Issue 7 December 2020
ISSN NO 2582-0958
________________________________________________________________________________________________

we get to know that concept of disese network is discovered a human disease network(HDN)
just on limited to disease but to various other where couple of disorders are connected to each
aspects like symptoms,treatment, precautions,etc . other if they have common genetic
code.Therefore Drug repurposing plays an
important role in disease networks and its
understanding, drug repurposing is also called as
drug repositioning where we utilize the medical
indications of previous drug and can use our
time ,cost and other factors for the disease that
are really needed to be cured.Traditional drug
development consumes alot of time ,energy and
cost therefore disease networks are the one that
help us to overcome the difficulties that we face
during inventing a drug. As we know that we
dont have the medication treatment for HIV
virus, but by giving them the treatment we can
prolong the life of a patient. Traditional drug
development As we know all the drugs do have
2. ACCESSING METHODOLY some side-effects , some diseases are acquired
by the environment while some are inherited
Due to alot of research was carried out on from genes, even if we try to cure a genetic
disease network and their contribution to disorder it has some or the other side effects,
understanding of drug and its repurposing we therefore we cannot cure genetic disorder by
found out that there were five databases traditional means of research. Genes are the
including Medline , Web of Science, Springer, segments of deoxyribonucleic acid(DNA) and
ACM digital library and IEEE Xplore. The riboxynucleic acid (RNA) which have 99.99%
results were confined only in English language similarity only 0.01% is the percentage which
studies published from 2007 to September are our own characteristics.
2008.
One of the research says that diseases are prone
Terms such as disease network, network to assemble by their classification and degree of
medicine, drug repurposing, pipeline and distribution that follows a power law,which
many more were used for the search. The means only few diseases are connected to a large
acquired studies and their results were number of diseases, whereas most diseases have
analysed by a single researcher and evaluated some links to others.In the year 2008, Lee et al
by various co-authors so that now it can be created a metabolic disease network where two
considered as the systematic validation. Based disorders were linked if the enzymes associated
on the eligibility criteria, there were analysis of with them catalyze adjacent reactions. In 2009, a
articles, the study was included in the complex disease gene network was created by
evaluation if: Barrenas et al using GWA( genome wide
(a) The studies inscribe the application of association), this network showed that disease
biological network to disease understanding which are under the same classification do not
and/or repurposing of drug. always share the common disease genes.
(b) Process to build a network. Therefore complex disease genes are primarily
(c)Also if qualitative and quantitative used compared to monogenic disease genes in
information of the generated network is human protein. The business application of drug
provided. repositioning and engrossment showed by
pharmaceutical companies have led to increase
3.MODIFICATION IN NETWORKS OF educational activity in this field.
DISEASES AND THEIR APPLICATION.
3.1.PRIOR STUDIES:
Human disease network was invented by Goh et
al , he created a disease-gene bipartition network
type of graph called “Diseasome” using
information from OMIM database, and they

www.trendytechjournals.com 36
 International Journal of Trendy Research in Engineering and Technology
Volume 4 Issue 7 December 2020
ISSN NO 2582-0958
________________________________________________________________________________________________

concluded that two diseases are connected to eac

other if the probablity of co-occouring in clinics
diverges from what was expected under self rule.
Despite their illustrated potential in pathological
analysis , the access and use of clinical record in
medical research is restricted by several issues
including heterogenity of sources, which are
ethical and legal.
An analysis was done by okumura et al in which
there was a plotting between clinical
vocabularies and findings in medical litreature
using OMIM as a knowledge source and
3.2 NETWORK MEDICINE metmap as NLP tool was done, following his
Through network we can discover many drugs design rodrigues et al found out diagnostic
and those drugs can be used for either clinical finding from medline plusarticles by
symptom based or genetic disease or disorder. using the web scraping and the combination of
Medicines can be made from networks through NLP techniques using metamap tool. In further
gene mapping, diagnose the symptom, or analysis the same team compared the
through prognostication. When we approach to performance of metamap and c in the same
a human disease through network, it becomes work.
easy for us to identify what kind of disease is it Zhou et al constructed a human symptom
or whether it has genotypic or phenotypic diseasse network (HDSN)using symptom
characters present in it. information from PubMed, in 2014. In HSDN
Rzhetsky et al took the initial step and started the influence between the two disease quantifies
his research in network disease using the the resemblence of their respective symptoms. In
phynotypic sources in 2007. The disease 2015, hoehndorf et al used a similarity measure
history of 1.5million peopleat columbia for text minded phenotypes and created another
university medical center concludes the human disease network.In both the cases the
comorbidity links between various kinds of disease share relationship with number of
disorder and prove that phenotypes set up a genetic union .They also indicate that not only
highly connected network of strong pairwise common ones are prone to be grouped into
relation.In 2009, Hidalgo et al raised classes but also the mendelian diseases.
phynotypic disease network (PDN) adding the
connections of more than 11 thousand disease 3.3.INTERACTIONS OF PROTEIN AND
obtained from a pairwise comorbidity relations DISEASOME
restored from over 30 million records from As we know diseasome means a group of
medicare patients. diseases and disorder including problems in
Actually phenotypic disease network(PDN) genetic code too. Due to the complexity arising
help in understanding the genesis of many between disease and disorders, the consideration
disease and their connection with each other. of a single factor whether it is a symtom disease,
or a gene disease or a drug disease was a
Phenotypic disease network are the network restricting factor to obtain new findings and
which affect one or more physiological system predicting the drug to be repositioned,
in a living beings. Phenotypic disease network understanding this concept of disease networks
is not aware to the mechanism which is various scientists came forward to give their
underlying the observed comorbidity but it contribution towards it, it is also called as
shows that patient tend to spread disease in the HDN(human disease network). In 2012, Goh et
network surrounding of disease they already al suggested that all the disease contributing
have. It is also found that succession of disease factor have to be integrated in a context
differs across genders and ethinicity. dependent manner for example molecular
linkage from protein interaction, co –
Recently an epidimeological HDN(eHDN) was expression, metabolism, genetic interactions and
created by jiang et al using data from taiwan phenotypic comorbidity links will be
national health insurance research data base and accordingly if we only consider them under
symptom or gene associated disease. But if the

www.trendytechjournals.com 37
 International Journal of Trendy Research in Engineering and Technology
Volume 4 Issue 7 December 2020
ISSN NO 2582-0958
________________________________________________________________________________________________

disease is drug based then, drug chemical called as DTI drug target indication. example
information, toxicity and non biological Rephetio , it was an activity based on hetionet
environmental factors will be considered. The which anticipate almost 3394 repurposing
result will be the integration of common applicant drugs by calculating the algorithm for
bipartition network which will be represented as social network analysis. Another meritorious
single, complex , k- partite heterogeneous example is DTInet where DTI prediction system
network , they are also called as complete is based on how low dimensional feature vectors
diseasome. After doing research Gottlieb found present a superior performance than other state
out the solution in the same lines he made a of art prediction methods and created a potential
wider collection of information sources and applied to COX inhibitors to prevent
created five drug drug similarity measure with inflammatory diseases.
two disease – disease similarity measures. It
means there will be interaction among five 4. DATA PIPELINE TO BUILD DISEASE
similar drugs with each other which will treat NETWORK
two similar diseases at one time. He used this As discussed in previous sections about
similarity measures for predicting calculations diseasome, and various types of disease
to find out novel drug which can be useful for networks, the research in this field has resulted
any kind of disease. Then scientists like Sun, in expanding the knowledge and trying to search
Albornoz, Daminelli, and Wang combined for more innovative ideas reguarding disease
various data sources and created tripartition networks. We always consider similarities
network of gene – disease – ppi , gene – diseease between diseases or we try to find out some of
– pathways, and drug – target – disease to the similar link when we want to connect one
predict which disease is associating with each disease to another, though we have aother
other and which drugs can be used for approches to find out disease networks for
repositioning. In 2012, heterogeneous networks example - computational approach or biological
are created from 17 different public data relating or experimental approach is also useful but we
to drugs, chemical compounds, protein targets, always try to find similarity among disease for
etc. This was done was Chen et al. Later Zitnik its repurposing, this is because the target organ
accelerated the research on disease networks by of a disease is same if they are related to each
incorporating ontology and molecular other so we have to block or stimulate that part
interactions by linking them to each other and of the receptor which can block the antigen to
created another heterogeneous network by enter the body. Both the drugs are having same
linking 11 different types of data on metaphysics target because they are arrived from same
and molecular bond interaction, but when this network, the symptoms may differ because
heterogeneous networks were getting evaluated antigen may attack different part of the receptor
one of the most notable points were genetic or due to some other biological issue.
interactions are most informative property as The function of the original drug and the
they are complicated and play their role in both repurposed drug differs by some points that is
chemistry and physics. In both the studies, why the drug is repurposed. Additionally the
logical thinking on phisophical networks as well drugs (both repurposed and original) share
as metaphsics in molecular bond were applied to common phases like modelling, visualization etc
illustrate the edges. A meritorious example is also can be represented as functional unit of data
Hetionet which is an integrative network born science.
after the studies of millions of billions of
biomedical research. Its data was combined from
29 different sources to connect disease,
pathways, biological processes and many more.
However the disadvantage was lack of standard
i.e a standard which was not established for
calibration and this may lead to uncertainity in
link description which may lead to misleading
interferences. As shown in fig1.
As there is advancement in technology, this has
resulted in the tools/ gadgets for the
prognostication of treatment of new disease

www.trendytechjournals.com 38
 International Journal of Trendy Research in Engineering and Technology
Volume 4 Issue 7 December 2020
ISSN NO 2582-0958
________________________________________________________________________________________________

4.1.DATA SCIENCE NETWORK AND ITS related to each other.

PROCESSING Diseases are conneted to each other when they
As there is growing availablity of information are connected to a particular gene or its part or if
due to technology and other traditional its a symptonm based disease then we can
methods there is improvement in the connect to both of them by finding out the
understanding of disease and its networks in hidden link or network between them that makes
various perspective. Data acquisition is them connected. Due to various
something which is the first and the initial incomprehensiblities arising in drug formulation
step, in data acquisition pipeline acquires data disease network has made it abit simple and
from different sources and present itself. In the appropriate.
second step pipeline consists of making the
work more appropriate by transforming and 4.2.1 SIMPLEST HOMOGENEOUS
mapping the given data. There are some NETWORKS
literature sources such as PubMed or GWAS They are the fine projections of heterogeneous
catalogue which is used for significant number networks, and are the most simplest one. Disease
of studies. Due to all these resources the - disease networks are usually validated by
standard of drugs and its application has using standard network method. In an
improved alot. Researchers use biological uncomplicated approach the link weights results
means of data like KEGG,Biogrid or OMIM in disease – disease network that represent link
along with its type and and description for its abundance from the previous projection data.
use in network study. Methods such as hyperbolic weighting and
Data science pipeline also processes action resource allocation weighting can be used as
towards reproduciblity and similarity or alternative though complex but are useful. As we
differences among studies as a whole or in know homogeneous networks are built on
phase level. Recent studies collect data from similarity networks. In these networks if the
books, experiments and various sources to give comparison of i and j takes place and if the
more accurate prediction capacity. This creates similarity points is greater than zero then the
challenges for identification of a proper corrosponding vertices are connected and a
network source, for all these solutions, network is created.
scientists used a word finder which addressed Homogeneous networks are the simplest
terms like MeSH, SNOMED ct or code listing. networks coming from human interactome and
Including all the logical and hierachial source, are based on various types like its composition,
we can allow maping data such as disease code duration of action, etc. The structure of protein
or any medical term. In medical literature influences function by determining molecules ith
sources Metadata is more often mixed up with which it interacts and outcome of interaction.
the word extraction tools such as Metamap.
5.CONCLUSION
4.2.DATA COMBINATION AND ITS At the end, we can say that network of diseases
STRUCTURING are complicated to identify but once we get also
To find out the answer to our question of our know the network we can easily find out the
study, we have to process the integrated data and disease and take precautions if possible. We can
analyse it properly. After this, networks of get to know the approach of virus or bacteria in
disease are constructed from the output data of it's particular way. It may or may not be
the previous stage and a model is made out of it. connected to genes but they are surely connected
As they are made of nodes and edges, they can to a network coming from other diseases.
or cannot be directed or weighted. Depending on
the type of node and its features they connect to REFERENCES
each other. Due to disease network, the
complexity in understanding the disease has led https://www.biorxiv.org/content/10.1101/4152
to various perturbations especially in human. In 57v3
data combination and structuring there arises https://www.sciencedirect.com/science/article/
there arise molecular complexity between pii/S1532046419301248
protein interaction and disease networks which https://en.m.wikipedia.org/wiki/Network_medi
become more stabilized when we know that the cine
structure of protein and the combined data are https://www.pnas.org/content/104/21/8685

www.trendytechjournals.com 39