Biomedical

Communication

Biosci. Biotech. Res. Comm. 10(1): 34-43 (2017)

Machine learning approaches to study HIV/AIDS

infection: A Review

Sweta Kumari, Usha Chouhan and Sunil Kumar Suryawanshi*

Department of Bioinformatics, Maulana Azad National Institute of Technology (MANIT), Bhopal- 462051

(M.P.), INDIA

ABSTRACT

In this review, PubMed database has been explored to elucidate the problems related to HIV/AIDS, which have

been solved previously using various machine learning approaches and some other techniques. Literatures from the

epidemic years of HIV/AIDS till February, 2017 have been examined and problems such as prediction of HIV/AIDS

protease cleavage sites and inhibitors, prediction of coreceptors usage for viral entry, development of anti-viral

agents and prediction of response, resistance and adverse effect of antiretroviral therapy have been considered for

the current study. Complications associated with HIV/AIDS infection as well as all three stages of HIV infection have

been described. HIV virus binding to the coreceptors CCR5 and CXCR4 are delineated to show the signi cant role of

the coreceptors for the anti-HIV drug development. After exploring various datasets, viral tropisms are found to be

relevant to the viral third V3 region of the HIV virus binding.

KEY WORDS: HIV, MACHINE LEARNING TECHNIQUES, V3 REGION, CD4 RECEPTOR, CCR5 AND CXCR4 CORECEPTORS

ARTICLE INFORMATION:

*Corresponding Author: bioinfosunil@gmail.com

Received 29

Dec, 2016

Accepted after revision 3

March, 2017

BBRC Print ISSN: 0974-6455

Online ISSN: 2321-4007 CODEN: USA BBRCBA

Thomson Reuters ISI ESC and Crossref Indexed Journal

NAAS Journal Score 2017: 4.31 Cosmos IF : 4.006

reserved.

Online Contents Available at: http//www.bbrc.in/

INTRODUCTION

Human immunode ciency virus/acquired immune de -

ciency syndrome (HIV/AIDS) was originated from mon-

key in the United States in 1981. AIDS is a chronic and

potentially most threatening infectious disease caused

by human immunode ciency virus in the 21

century.

78 million people were estimated to be suffering from

HIV/AIDS and 35 million people have died since the

start of the epidemic year but 36.7 million people were

reported as HIV/AIDS infected and 1.1 million people

have died in 2015 globally; 2.1 million people were

found to be newly HIV/AIDS infected globally (Alkema

2016). Eastern and Southern Africa have a maximum

increase since almost the start of the epidemic year. HIV/

AIDS regularly decimated the population of the Africa

shown in  gure1 (HIV/AIDS 2016).

The current prevalence of the HIV/AIDS is 0.8% of

the worldwide. 18.2 million people were accessing

antiretroviral therapy in June 2016 (Alkema et al. 2016).

Sweta, Usha and Sunil

Antiretroviral therapy (ART) has a very crucial role in

the HIV treatment, but HIV infected individuals acquire

resistance to the ART over a certain period. Antiretrovi-

ral therapy coverage increased from 7.51 million in 2010

to 17.03 million in 2015 shown in  gure 2.

However, scientists are working to develop one vac-

cine to prevent HIV/AIDS. NIH supported clinical trial

that was launched in 2016 to test a possible HIV/AIDS

vaccine. This vaccine trial HVTN 072 is testing whether an

experimental vaccine regimen safely prevents HIV/AIDS

infection among South African adults (Health 2015).

The determination of the 90-90-90 treatment target

is to be achievable by the reinforcement of the con-

tinuing momentum in 2020, whereby 90% of the HIV/

AIDS infected people aware of their HIV/AIDS status,

90 % of the people with known their HIV/AIDS positive

status are taking treatment and 90% of the people on

HIV/AIDS treatment have suppressed viral loads (NAS-

COP 2014). Recent updated UNAIDS estimates indicate

that US$ 26.2 billion will be required for the HIV/AIDS

response in 2020, with US$ 23.9 billion required in 2030

(NASCOP 2014). The world has committed to ending the

HIV/AIDS epidemic by 2030 (Bernard 2016).

METHODS

We searched PubMed database with the keywords of

“HIV” and “Machine Learning Approaches” up to Feb-

ruary 22, 2017 and starting time was not given. The

number of the articles retrieved from the PubMed data-

base was 114, out of which there was 6 review articles.

Efforts here made to review some uncertainties related

to HIV infection. The search completeness was examined

by the list of references of the reviewed articles. Some-

where, Book chapters were also included, some articles

were excluded on criteria such as articles not in English

language, campaign posters, newspaper articles and the

articles irrelevant to the topic of this review. In numbers,

32 originally related articles were selected on the basis

of this evaluation are listed in Table-1.

FIGURE 1. New HIV infections among people aged 15 years and above in different

regions since last six years from 2010 to 2015(Alkema et al. 2016)

FIGURE 2. Antiretroviral therapy coverage % in different regions since last six

years from 2010 to 2015(Bernard 2016)

BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION 35

Sweta, Usha and Sunil

STAGES OF HIV/AID

There are mainly three stages of HIV: Primary infection

(Acute HIV), Clinical latent infection (Chronic HIV) and

Early symptomatic HIV infection (Wawer 2005). In pri-

mary infection, a  u-like illness developed within one or

two months after the virus entering the body of people,

including signs and symptoms such as fever, headache,

muscle aches, joint pain, rash, sore throat and swollen

lymph glands mainly on the neck. Though, the symp-

toms of the  rst stage of HIV infection are mostly unno-

ticed, the amount of virus in the viral load or bloodstream

spreads very highly at this time, resulting in dispersing

the HIV infection more ef ciently during primary infec-

tion than the next stage. In the clinical latent infection,

generally persistent swelling of lymph nodes occurs and

HIV remains in the human body but with no signs and

symptoms. This stage lasts around 10 years in people not

receiving antiretroviral medications and lasts for decades

in people receiving antiretroviral therapy. Some people

expedite to more intense disease much sooner. In the last

stage of early symptomatic HIV infection, people probably

progress to mild infections or chronic signs and symp-

toms such as fever, fatigue, swollen lymph nodes, diarrhea,

weight loss, oral yeast infection (thrush) and shingles (her-

pes zoster) due to continuous replication of the virus in the

human body and distortion of the immune cells.

COMPLICATIONS OF HIV/AIDS

The burden of HIV is to a large extent consequence of

infections (tuberculosis (TB), cytomegalovirus, candidi-

asis, cryptococcal meningitis, toxoplasmosis and crypto-

sporidiosis), cancers (Kaposi’s sarcoma and lymphomas)

and others (wasting syndrome, neurological complica-

tions and kidney dis How ease) (How). The increasing in

the incidences of the HIV/AIDS and their consequences

in terms of falling down the number of CD4 (Cluster of

differentiation 4) receptors and coreceptors lead to dam-

age the immune system of the human. Therefore, iden-

ti cation of the used coreceptor for the viral entry that

can target to block the coreceptor to bind with the virus

and maintenance of the CD4 receptors and coreceptors

is being crucial for new therapeutic agents for the treat-

ment of immune de ciency syndrome (Kaplan 2009).

Infections common to HIV/AIDS

Tuberculosis (TB): Tuberculosis coinfection is associated

with increase viral replications. TB is the most occurring

infection related to HIV/AIDS and a leading cause of death

(Pawlowski 2012).

Cytomegalovirus: This herpes virus is transmitted in

human body  uids such as saliva, urine, blood, breast milk

and semen. The virus remains inactive in a healthy immune

system and remains resurfaces in weakens immune sys-

tem causing damage to eyes, lungs, digestive tract or other

organs (Mathevula 2013).

Candidiasis: Candidiasis causes in ammation and a thick,

white coating on the mucous membranes of mouth, tongue,

vagina or esophagus (Cutlan 2010).

Cryptococcal Meningitis: Cryptococcal meningitis is a

common infection of the central nervous system associated

with HIV/AIDS, caused by a fungus found in soil. Menin-

gitis is an in ammation of the membranes and  uid sur-

rounding brain and spinal cord (Park 2009).

Toxoplasmosis: This infection is caused by Toxoplasma

gondii, a parasite dispersed primarily by cats. Infected

cats successfully pass the parasites in their stools, and the

parasites may then transfer to other animals and humans

(Berger-Schoch 2011).

Cryptosporidiosis: An intestinal parasite usually found

in animals and it is responsible for the infection. Crypto-

sporidiosis contracted when human ingests contaminated

food or water. The parasite grows in the intestines of human

and bile ducts, leading to severe, chronic diarrhea in HIV/

AIDS infected people (Mathewos 2014).

Cancers common to HIV/AIDS

Kaposi’s sarcoma: Kaposi’s sarcoma is a blood vessel walls,

a very rare tumor in HIV-negative people but a very com-

mon in HIV-positive people. It usually appears as pink, pur-

ple or red lesions on the skin and mouth of the HIV infected

people (Page 2006). Internal organs can also be affected by

Kaposi’s sarcoma (Di Benedetto 2008).

Lymphomas: Lymphomas cancer originates in white blood

cells and usually appears in lymph nodes. Painless swelling

of the lymph nodes in neck, armpit or groin is most com-

mon early sign (Daniels 2012).

Other complications

Wasting syndrome: Aggressive treatment regimens have

decreased the number of cases of wasting syndrome, but

it still affects many HIV/AIDS infected people. It leads to

a loss of at least 10 percent of human body weight, often

escorted by diarrhea, chronic weakness and fever (Bass

2015).

Neurological complications: It can cause neurological

symptoms such as confusion, depression, forgetfulness,

anxiety and dif culty walking, whereas AIDS doesn’t

infect the nerve cells. AIDS dementia complex is one of

the most common neurological complications, which leads

to changes in behavior and diminishes in mental function

(Ances 2007).

Kidney disease: HIV-associated nephropathy (HIVAN) is an

in ammation of the small  lters in kidneys that remove

extra  uid and wastes from bloodstream and pass them to

your urine. The risk of developing HIVAN is much higher

in black people due to a genetic predisposition. A antiret-

roviral therapy should be started in those diagnosed with

HIVAN in any case of CD4 count (Scherzer 2012).

36 MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS

Sweta, Usha and Sunil

Receptors

CD4 receptors are particularly attractive target mol-

ecules, since they have been found to play a certain role

in maintaining the immune system of the human. CCR5

and CXCR4 are two coreceptors of the receptor present

on the T-cell, which is useful for viral cellular entry

shown in  gure 3.

CCR5 (C-C chemokine receptor type 5)

The CCR5 protein belongs to the family of the beta

chemokine receptors of the integral membrane proteins

(Samson 1996). The CCR5 coreceptors are G protein-

coupled coreceptors (Aavikko 2014). CCR5 coreceptors

are also known as CD195 proteins expressed on the sur-

face of white blood cells, T helper cells, macrophages,

dendritic cells, eosinophils and microglia. T helper cells

are speci c tissues and organ targets by which HIV virus

causes AIDS using CCR5 coreceptors to enter and infect

into the immunological cells. The viral entry of HIV-1

into a target host cell is enabled by the essential HIV-1

envelope glycoprotein structure (Alkhatib 2009). The

Gp120 external subunit and Gp41 transmembrane subu-

nit are two subunits of the envelope glycoprotein struc-

ture cleaved from a Gp160 protein precursor encoded by

the HIV-1 env gene (Checkley 2011).

The Gp120 subunit binds to a CD4 glycoprotein and

a HIV-1 coreceptor CCR5 expressed on a target host cell

forming a heterotrimeric complex (Murphy 2001). The

binding of Gp120 envelope protein to CCR5 coreceptor

consists two crucial steps: The tyrosine sulfated amino

terminus of this coreceptor is an “essential determinant”

of binding to Gp120 in the  rst step and there must be

reciprocal action (synergy, intercommunication) between

gp120 and the CCR5 transmembrane domains in a sec-

ond step (Ji 2007). Some individuals lead to a muta-

tion as CCR5-delta 32 in the CCR5 gene by the genetic

deletion of a portion of the CCR5 gene to protect them

against these HIV strains. This mutation in homozygous

carriers is resistant to the M-tropic strains of the HIV-1

infected individuals (De Silva 2004 Hütter 2009 Allers

2011 Zhen 2013 Kay 2014 and Tebas 2014).

The CCR5 gene encodes the CCR5 protein, which is

located on the short (p) arm at the position of 21 on chro-

mosome 3.The cognate ligands of CCR5 include CCL3,

CCL4 and CCL3L1 (Struyf 2001 and Miyakawa 2002) and

CCR5 interacts with CCL5 (Struyf et al. 2001 Miyakawa

et al. 2002 and Slimani 2003).The general ligands for the

receptor RANTES, MIP-1 and MIP-1 are able to sup-

press HIV-1 infection in vitro (Nakayama 2012).

CXCR4 (C-X-C chemokine receptor type 4)

CXCR4 is an alpha-chemokine receptor speci c for

stromal-derived-factor-1 (SDF-1) also known as CD184

(cluster of differentiation 184) proteins. SDF-1 is a mol-

ecule endowed with effective chemotactic activity of

lymphocytes. During the implantation and hormone

replacement therapy cycle in the endomentrium, pro-

ducing, in the presence of human blastocyst and a sur-

face polarization of the CXCR4 coreceptors suggest that

these coreceptors are connected in the adhesion phase

of the human implantation, CXCR4 is upregulated. The

ligand of CXCR4, SDF-1 is known to be signi cant in

hematopoietic stem cell homing to the bone marrow and

in hematopoietic stem cell quiescence. CXCR4 signaling

coreceptors found to regulate the CD20 expression on B

cells (Pavlasova 2016).

SDF-1 and CXCR4 were believed to be a relatively

monogamous ligand-receptor pair (other chemokines are

promiscuous, tending to use several different chemokine

receptors) until recently. Recent evidence demonstrates

ubiquitin is also a natural ligand of CXCR4 (Saini 2010).

Ubiquitin is a small (76-amino acid) protein highly con-

served among eukaryotic cells. It is best known for its

intracellular role in targeting ubiquitylated proteins for

degradation via the ubiquitin proteasome system. Evi-

dence in numerous animal models suggests ubiquitin is

anti-in ammatory immune modulator and proin am-

matory damage endogenous opponent associated molec-

ular pattern molecules (Majetschak 2011). It is speculated

this interaction may be through CXCR4 mediated sign-

aling pathways. MIF is an additional ligand of CXCR4

(Bernhagen 2007).

CXCR4 plays a role in neurological guidance by pre-

senting in newly generated neurons during embryo-

genesis and adult life. The receptor levels decrease as

neurons mature. CXCR4 mutant mice have aberrant

neuronal distribution. This has been implicated in disor-

ders such as epilepsy (Bagri 2002).

The viral regions

Coreceptor consideration holds the V1-V2 region

oh Gp120 and the bridging sheet such as antiparallel

and 4-stranded  sheet that joints the inner and outer

domains of the Gp120. The coreceptor usage through its

peptide composition and the degree of N-linked glyco-

sylation can be in uenced by the V1-V2 stem. In con-

trast with V1-V2 region, the V3 loop region is highly

mutable and therefore it is the most signi cant determi-

nant of coreceptor speci city (Bozek 2013).

The HIV enzymes role

The viral genome reverse transcription is required to

generate the proviral DNA and the integration into the

target genome cell for successful HIV-1 virus replication.

The viral encoded enzymes such as reverse transcriptase

(RT) and integrase (IN) put their potential for these

events and act sequentially during viral replication. The

BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION 37

Sweta, Usha and Sunil

FIGURE 3. HIV virus cellular entry through binding with the CD4 receptor and

CXCR4 and CCR5 coreceptor in T-tropic and M-tropic HIV respectively(Gorry

2005)

double stranded proviral DNA produced by replicating

both RNA and DNA templates using RT enzyme which

is a heterodimer of p66 and p51 subunits (Arts 1997).

A de ned set of DNA cleavage and combined events to

insert the proviral DNA into the host genome directed by

the IN enzyme which is 32 kDa polypeptide (Wiskerchen

1995). HIV-1 protease (PR) is important for the life-

cycle of the HIV virus and this enzyme is responsible

for cleavage of the newly synthesized polyproteins

at the appropriate places to create the mature protein

components of an infectious HIV virion. HIV virions

remain uninfected without effective HIV PR (Doitsh

2010).

The review highlights the scarcity of information on

the HIV/AIDS problems which are possible to solve using

various machine learning techniques are shortlisted in

Table 1. Classi cation of X4, R5 and dual (R5X4) tropic

HIV-1 from V3 loop amino acid sequences of HIV-1 sub-

type B using arti cial neural networks up to accuracy of

81.8% (Fogel et al. 2015) and classi cation of X4 and

R5tropic HIV-1 from V3 loop amino acid sequences of

HIV-1 subtype C using SVM showed a good concord-

ance of 85% (Gupta et al. 2015) are done in USA and

India respectively. Prediction of coreceptor usage for

viral entry from Gp120 V3 loop amino acid sequences

using SVM, heuristic and statistical learning methods,

two-level machine, random forest, boosted decision

tree, and neural network machine learning algorithms

has been done in India, Germany and USA respectively

(Raghava 2013 Sander et al. 2007 Dybowski 2010 and

Vaisman 2010). Prediction of HIV-1 protease cleavage

site and inhibitors using Feature selection subset method

of multi-layered perceptron (FS-MLP) learning, SVM,

ANN pharmacophore and docking methods have been

done 80.0% ~ 97.4% accuracy in Taiwan, Korea, Turkey

and China respectively (Kim et al. 2010 Singh Su 2016

Öztürk et al. 2013 and Wei et al. 2015).

Some of the literatures are regarding of drug devel-

opment, anti-viral agents development (Kirchmair et

al. 2011), antiretroviral response prediction (Zazzi et al.

2012 Prosperi 2011 and Prosperi et al. 2009), antiretro-

viral resistance prediction (Zazzi 2016 Riemenschneider

2016a Riemenschneider 2016b Heider et al. 2013 and

Kijsirikul 2008), antiretroviral adverse effects predic-

tion (Adrover et al. 2015) has been analyzed which were

used machine learning approaches, SVM, Expert’s Rules

and Linear and Non-linear statistical learning algo-

rithms, Radial basis function networks (RBF networks),

k-nearest neighbor (kNN) and Virtual screening method

to improve their result in Austria, Italy, USA, Germany

and Thailand. Prediction of antibody of HIV epitope net-

works using neutralization titers and a novel computa-

tional methods or a simple machine learning methods

has been done in USA (Evans et al. 2014 Hepler et al.

2014 and Choi et al. 2015.

Prediction of HIV-1 RT associated RNase H inhibition

(Poongavanam 2013) shown good enrichment (80-90%)

by receptor-based  exible docking experiments com-

pared to ligand-based approaches such as FLAP (74%),

shape similarity (75%) and random forest (72%) in Den-

mark. Ligand based computational modeling studies on

non-nucleoside reverse transcriptase inhibitors of HIV-1

(Pancholi et al. 2014) has been done India respectively.

Prediction of bioactivities of HIV-1 integrase ST inhibi-

tors (Xuan et al. 2013) and classi cation of active and

weakly active ST inhibitors of HIV-1 integrase has (Yan

et al. 2012) been done using machine learning approaches

in China and USA respectively. Prediction of interactions

between HIV-1 and human proteins using SVM in China

has also done (Wei et al. 2015). One of the literatures has

worked on the detection of M tuberculosis in patients with

and without HIV coinfection by identifying a 251 gene

expression signatures using SVM with 81.4% and 88.9-

94.7% accuracy respectively in USA (Dawany et al. 2014).

38 MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS

Sweta, Usha and Sunil

Table 1. Literatures involved for the current study for machine learning approaches used to solve various HIV/

AIDS problems.

Number Author name-

Publication year

Methods Objectives Country

1 (Gupta 2015) Machine learning

methods

Prediction of tropism in HIV-1 subtype C V3 loop

sequences using genotypic tools.

India

2 (Fogel 2015) Arti cial neural

networks

Classi cation of R5-, X4- and R5X4-tropics HIV-1

using evolved neural networks.

3 (Antell 2016) Machine learning Identi cation of R5- and X4-speci c Tat and LTR

sequence signatures using HIV envelope V3.

USA

4 (Kumar 2013) Support vector

machine

Prediction of HIV-1 coreceptor usages using hybrid

approach from its V3 loop amino acid sequences.

India

5 (Sander 2007) Heuristic and

Statistical learning

methods

Prediction of HIV-1 coreceptor usage by structural

descriptors of gp120 V3 loop.

Germany

6 (Dybowski 2010) Two-level machine

learning methods

Prediction of co-receptor usage of HIV-1 from

genotype.

Germany

7 (Masso 2010) Random forest,

support vector

machine, boosted

decision tree, and

neural network

machine learning

algorithms

Determination of HIV-1 co-receptor usage from

accurate and ef cient gp120 V3 loop structure based

models.

USA

8 (Evans 2013) Support vector

machine, PSSM and

11/25 rule

Genotypic prediction of HIV-1 coreceptor tropism

using a case-based reasoning system.

USA

9 (Kim 2010) Feature selection

subset method

of multi-layered

perceptron (FS-MLP)

learning

Analysis of HIV-1 protease cleavage site an MLP-

based feature subset selection.

Korea

10 (Singh 2016) Arti cial neural

networks

HIV-1 protease cleavage site prediction using

a combination of sequence, structural, and

physiochemical properties.

Taiwan

11 (Öztürk 2013) Support vector

machine

Prediction of HIV-1 protease cleavage site using a

consistency-based feature selection method allied

with linear SVMs.

Turkey

12 (Wei 2015) Support vector

machine, shape,

pharmacophore and

docking methods

Multistage virtual screening and novel identi cation

of HIV-1 protease inhibitors by integrating SVM,

shape, pharmacophore and docking methods.

China

13 (Kirchmair 2011) Virtual screening

techniques.

Development of anti-viral agents using molecular

modeling and virtual screening techniques.

Austria

14 (Zazzi 2012) Machine learning

approaches

Prediction of antiretroviral treatment response using

machine learning: The EuResist project.

Italy

15 (Prosperi 2011) Machine learning

approaches

Prediction of antiretroviral treatment response using

computational models.

USA

16 (Prosperi 2009) Linear and Non-linear

statistical learning

models

Prediction of response to antiretroviral treatment by

investigating of expert rule bases, logistic regression,

and non-linear machine learning techniques.

Italy

17 (Zazzi 2016) Expert’s Rules and

Statistical/Machine

learning algorithms.

Computer-Aided Optimization of Combined Anti-

Retroviral Treatment for HIV: New Drugs, Drug

Targets and Drug Resistance.

BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION 39

Sweta, Usha and Sunil

18 (Riemenschneider

2016a)

Support vector

machine, Random

forest and Statistical

methods

Prediction of HIV Drug Resistance using Current

Computational Approaches.

Germany

19 (Riemenschneider

2016b)

Machine learning

techniques, binary

relevance classi ers,

classi er chains, and

ensembles of classi er

chains

Multiclass classi cation for HIV-1 drug resistance

prediction by exploiting cross-resistance information

with 662 protease sequences and 715 reverse

transcriptase sequences.

Germany

20 (Heider 2013) Machine learning

techniques.

Multiclass classi cation for HIV-1 drug resistance

prediction by exploiting cross-resistance information.

Germany

21 (Srisawat 2008) Support vector

machine, Radial basis

function Networks and

k-nearest neighbor

methods

Prediction of HIV-1 drug resistance by combining

classi ers.

Thailand

22 (Adrover et al.

2015)

Machine learning and

crowdsourced human

assessment

Identi cation of adverse effects of HIV drug

treatment and related sentiments using Twitter.

23 (Evans 2014) Boosted algorithms of

consisting of multiple

machine learnings and

statistical models

Prediction of HIV-1 broadly neutralizing antibody

epitope networks using neutralization titers and a

novel computational method.

USA

24 (Hepler 2014) Machine learning

methods

IDEPI: fast prediction of HIV-1 antibody epitopes

and other phenotypic features from sequence dataset

using a  exible machine learning techniques.

USA

25 (Choi 2015) Machine learning

methods

Prediction of antibody feature: function relationships

in RV144 vaccines using machine learning methods.

USA

26 (Poongavanam

2013)

Virtual screening Prediction of HIV-1 RT related to RNase H inhibition

using virtual screening.

Denmark

27 (Pancholi 2014) SVM, Back

propagation neural

networks

Ligand based computational modeling studies on

non-nucleoside reverse transcriptase inhibitors of

HIV-1

India

28 (Xuan 2013) Support vector

machine and

Regression methods

Bioactivity of HIV-1 integrase ST inhibitors predicted

using multilinear regression analysis and support

vector machine.

China

29 (Yan 2012) Machine learning

methods

Support vector machine used for classi cation of

active and weakly active ST inhibitors of HIV-1

integrase.

USA

30 (Mei 2013) Support vector

machine

Prediction of interactions between HIV-1 and human

proteins using probability weighted ensemble transfer

learning.

China

31 (Dawany 2014) Support vector

machine

Detection of accurate M. tuberculosis in patients with

and without HIV co-infection by identi cation of a

251 gene expression signature.

USA

32 (Holman 2012) Machine learning

methods

Identi cation of amino acid signatures in the HIV

env gene predictive of dementia using a machine

learning approach.

USA

CONCLUSION

After reviewing all the literature, the coreceptor which

has been used for cellular viral entry is very necessary

to identify to develop the drugs that can target the core-

ceptor and prevent the coreceptor from binding with

HIV virus. Maraviroc has been found to be an as suc-

cessful barrier in blocking the CCR5 coreceptor but fails

in blocking the CXCR4 coreceptor. Prediction of HIV-1

coreceptor usage is necessary to identify the number of

CD4 counts remained into the body of HIV-1 infected

people. HIV-1 protease is a retroviral aspartyl protease

(retropepsin) that is essential for the HIV-1 lifecycle,

the retrovirus that is responsible for AIDS. The integral

40 MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS

Sweta, Usha and Sunil

role of the HIV-1 protease in the viral replication, HIV-1

protease has become a prime target for drug therapy.

Identi cation of the coreceptor which has been used

for cellular viral entry with different machine learning

approaches and more high accuracy than the previous

 ndings is our future work for contribution in the drug

development of HIV-1.

ACKNOWLEDGEMENT

We are thankful to the Department of Biotechnology

(DBT), New Delhi for providing support for this work

under Bioinformatics Infrastructure Facility of DBT at

Maulana Azad National Institute Technology, Bhopal.

REFERENCES

Aavikko, M. (2014) Identi cation of novel tumor predisposi-

tion families and underlying genetic defects.

Adrover, C., Bodnar, T., Huang, Z., Telenti, A. ,Salathé, M.

(2015) Identifying adverse effects of HIV drug treatment and

associated sentiments using Twitter. JMIR public health and

surveillance, Vol. 1, No. 2.

Alkema, L., Chou, D., Hogan, D., Zhang, S., Moller, A.-B.,

Gemmill, A., Fat, D.M., Boerma, T., Temmerman, M. ,Mathers,

C. (2016) Global, regional, and national levels and trends in

maternal mortality between 1990 and 2015, with scenario-

based projections to 2030: a systematic analysis by the UN

Maternal Mortality Estimation Inter-Agency Group. The Lan-

cet, Vol. 387, No. 10017, pp. 462-74.

Alkhatib, G. (2009) The biology of CCR5 and CXCR4. Current

Opinion in HIV and AIDS, Vol. 4, No. 2, p. 96.

Allers, K., Hütter, G., Hofmann, J., Loddenkemper, C., Rieger,

K., Thiel, E. ,Schneider, T. (2011) Evidence for the cure of HIV

infection by CCR5Δ32/Δ32 stem cell transplantation. Blood,

Vol. 117, No. 10, pp. 2791-9.

Ances, B.M. ,Ellis, R.J. (2007) Dementia and neurocognitive

disorders due to HIV-1 infection, Seminars in neurology, Vol.

Seventh Avenue, New York, NY 10001, USA., pp. 086-92.

Antell, G.C., Dampier, W., Aiamkitsumrit, B., Nonnemacher,

M.R., Jacobson, J.M., Pirrone, V., Zhong, W., Kercher, K., Pas-

sic, S. ,Williams, J.W. (2016) Utilization of HIV-1 envelope V3

to identify X4-and R5-speci c Tat and LTR sequence signa-

tures. Retrovirology, Vol. 13, No. 1, p. 32.

Arts, E.J., Le Grice, S.F. (1997) Interaction of retroviral reverse

transcriptase with template–primer duplexes during replica-

tion. Progress in nucleic acid research and molecular biology,

Vol. 58, pp. 339-93.

Bagri, A., Gurney, T., He, X., Zou, Y.-R., Littman, D.R., Tessier-

Lavigne, M. ,Pleasure, S.J. (2002) The chemokine SDF1 reg-

ulates migration of dentate granule cells. Development, Vol.

129, No. 18, pp. 4249-60.

Bass, L.H., Washington, C.M. (2015) Infection Control in Radia-

tion Oncology Facilities. Principles and Practice of Radiation

Therapy, p. 178.

Berger-Schoch, A., Herrmann, D., Schares, G., Müller, N., Ber-

net, D., Gottstein, B. ,Frey, C. (2011) Prevalence and genotypes

of Toxoplasma gondii in feline faeces (oocysts) and meat from

sheep, cattle and pigs in Switzerland. Veterinary parasitology,

Vol. 177, No. 3, pp. 290-7.

Bernard, E., Cameron, S. (2016) Advancing HIV justice 2.

Building momentum in global advocacy against HIV crimi-

nalisation.

Bernhagen, J., Krohn, R., Lue, H., Gregory, J.L., Zernecke, A.,

Koenen, R.R., Dewor, M., Georgiev, I., Schober, A. ,Leng, L.

(2007) MIF is a noncognate ligand of CXC chemokine recep-

tors in in ammatory and atherogenic cell recruitment. Nature

medicine, Vol. 13, No. 5, pp. 587-96.

Bozek, K., Lengauer, T., Sierra, S., Kaiser, R. ,Domingues, F.S.

(2013) Analysis of physicochemical and structural properties

determining HIV-1 coreceptor usage. PLOS Comput Biol, Vol.

9, No. 3, p. e1002977.

Checkley, M.A., Luttge, B.G. ,Freed, E.O. (2011) HIV-1 enve-

lope glycoprotein biosynthesis, traf cking, and incorporation.

Journal of molecular biology, Vol. 410, No. 4, pp. 582-608.

Choi, I., Chung, A.W., Suscovich, T.J., Rerks-Ngarm, S., Pit-

isuttithum, P., Nitayaphan, S., Kaewkungwal, J., O’Connell,

R.J., Francis, D. ,Robb, M.L. (2015) Machine learning methods

enable predictive modeling of antibody feature: function rela-

tionships in RV144 vaccinees. PLOS Comput Biol, Vol. 11, No.

4, p. e1004185.

Cutlan, J., Saunders, N., Olsen, S. ,Fullen, D. (2010) White sponge

nevus presenting as genital lesions in a 28‐year‐old female.

Journal of cutaneous pathology, Vol. 37, No. 3, pp. 386-9.

Daniels, V.G. (2012) AIDS: the acquired immune de ciency

syndrome, Springer Science & Business Media.

Dawany, N., Showe, L.C., Kossenkov, A.V., Chang, C., Ive, P.,

Conradie, F., Stevens, W., Sanne, I., Azzoni, L. ,Montaner, L.J.

(2014) Identi cation of a 251 gene expression signature that

can accurately detect M. tuberculosis in patients with and

without HIV co-infection. PloS one, Vol. 9, No. 2, p. e89925.

De Silva, E. ,Stumpf, M.P. (2004) HIV and the CCR5-Δ32 resist-

ance allele. FEMS microbiology letters, Vol. 241, No. 1, pp. 1-12.

Di Benedetto, F., Di Sandro, S., De Ruvo, N., Berretta, M.,

Masetti, M., Montalti, R., Ballarin, R., Cocchi, S., Potenza, L.

,Luppi, M. (2008) Kaposi’s sarcoma after liver transplantation.

Journal of cancer research and clinical oncology, Vol. 134, No.

6, pp. 653-8.

Doitsh, G., Cavrois, M., Lassen, K.G., Zepeda, O., Yang, Z., San-

tiago, M.L., Hebbeler, A.M. ,Greene, W.C. (2010) Abortive HIV

infection mediates CD4 T cell depletion and in ammation in

human lymphoid tissue. Cell, Vol. 143, No. 5, pp. 789-801.

Dybowski, J.N., Heider, D. ,Hoffmann, D. (2010) Prediction of

co-receptor usage of HIV-1 from genotype. PLOS Comput Biol,

Vol. 6, No. 4, p. e1000743.

BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION 41

Sweta, Usha and Sunil

Evans, M.C., Paquet, A.C., Huang, W., Napolitano, L., Frant-

zell, A., Toma, J., Stawiski, E.W., Goetz, M.B., Petropoulos,

C.J. ,Whitcomb, J. (2013) A case-based reasoning system for

genotypic prediction of hiv-1 co-receptor tropism. Journal of

bioinformatics and computational biology, Vol. 11, No. 04, p.

1350006.

Evans, M.C., Phung, P., Paquet, A.C., Parikh, A., Petropoulos,

C.J., Wrin, T. ,Haddad, M. (2014) Predicting HIV-1 broadly

neutralizing antibody epitope networks using neutralization

titers and a novel computational method. Bmc Bioinformatics,

Vol. 15, No. 1, p. 77.

Fogel, G.B., Lamers, S.L., Liu, E.S., Salemi, M. ,McGrath, M.S.

(2015) Identi cation of dual-tropic HIV-1 using evolved neural

networks. Biosystems, Vol. 137, pp. 12-9.

Gorry, P.R., Churchill, M., Crowe, S.M., Cunningham, A.L.

,Gabuzda, D. (2005) Pathogenesis of macrophage tropic HIV-1.

Current HIV research, Vol. 3, No. 1, pp. 53-60.

Gupta, S., Neogi, U., Srinivasa, H. ,Shet, A. (2015) Performance

of genotypic tools for prediction of tropism in HIV-1 subtype C

V3 loop sequences. Intervirology, Vol. 58, No. 1, pp. 1-5.

Heider, D., Senge, R., Cheng, W. ,Hüllermeier, E. (2013) Multi-

label classi cation for exploiting cross-resistance information

in HIV-1 drug resistance prediction. Bioinformatics, Vol. 29,

No. 16, pp. 1946-52.

Hepler, N.L., Schef er, K., Weaver, S., Murrell, B., Richman,

D.D., Burton, D.R., Poignard, P., Smith, D.M. ,Pond, S.L.K.

(2014) IDEPI: rapid prediction of HIV-1 antibody epitopes and

other phenotypic features from sequence data using a  exible

machine learning platform. PLOS Comput Biol, Vol. 10, No. 9,

p. e1003842.

Holman, A.G. ,Gabuzda, D. (2012) A machine learning approach

for identifying amino acid signatures in the HIV env gene pre-

dictive of dementia. PloS one, Vol. 7, No. 11, p. e49538.

Hütter, G., Nowak, D., Mossner, M., Ganepola, S., Müßig, A.,

Allers, K., Schneider, T., Hofmann, J., Kücherer, C. ,Blau, O.

(2009) Long-term control of HIV by CCR5 Delta32/Delta32

stem-cell transplantation. New England Journal of Medicine,

Vol. 360, No. 7, pp. 692-8.

Ji, C., Zhang, J., Dioszegi, M., Chiu, S., Rao, E., Cammack, N.,

Brandt, M. ,Sankuratri, S. (2007) CCR5 small-molecule antago-

nists and monoclonal antibodies exert potent synergistic anti-

viral effects by cobinding to the receptor. Molecular pharma-

cology, Vol. 72, No. 1, pp. 18-28.

Kaplan, J.E., Benson, C., Holmes, K.K., Brooks, J.T., Pau, A.,

Masur, H., Control, C.f.D., Prevention, Health, N.I.o. ,America,

H.M.A.o.t.I.D.S.o. (2009) Guidelines for prevention and treat-

ment of opportunistic infections in HIV-infected adults and

adolescents. MMWR Recomm Rep, Vol. 58, No. RR-4, pp.

1-207.

Kay, M.A., Walker, B.D. (2014) Engineering cellular resistance

to HIV. N Engl J Med, Vol. 370, No. 10, pp. 968-9.

Kim, G., Kim, Y., Lim, H. ,Kim, H. (2010) An MLP-based feature

subset selection for HIV-1 protease cleavage site analysis. Arti-

 cial intelligence in medicine, Vol. 48, No. 2, pp. 83-9.

Kirchmair, J., Distinto, S., Roman Liedl, K., Markt, P., Maria

Rollinger, J., Schuster, D., Maria Spitzer, G. ,Wolber, G. (2011)

Development of anti-viral agents using molecular modeling

and virtual screening techniques. Infectious Disorders-Drug

Targets (Formerly Current Drug Targets-Infectious Disorders),

Vol. 11, No. 1, pp. 64-93.

Kumar, R., Raghava, G.P. (2013) Hybrid approach for predict-

ing coreceptor used by HIV-1 from its V3 loop amino acid

sequence. PloS one, Vol. 8, No. 4, p. e61437.

Majetschak, M. (2011) Extracellular ubiquitin: immune modu-

lator and endogenous opponent of damage-associated molecu-

lar pattern molecules. Journal of leukocyte biology, Vol. 89,

No. 2, pp. 205-19.

Masso, M., Vaisman, I.I. (2010) Accurate and ef cient gp120

V3 loop structure based models for the determination of HIV-1

co-receptor usage. Bmc Bioinformatics, Vol. 11, No. 1, p. 494.

Mathevula, H.M. (2013) Factors affecting adherence to treat-

ment in patients on chronic medication at Mokopane Hospital,

University of Limpopo (Tur oop Campus).

Mathewos, M. (2014) Prevalence of opportunistic intestinal

parasitic infections among hiv/aids patients attending Othona

hospital, Wolayita Sodo, Southern Ethiopia.

Mei, S. (2013) Probability weighted ensemble transfer learning

for predicting interactions between HIV-1 and human proteins.

PloS one, Vol. 8, No. 11, p. e79606.

Miyakawa, T., Obaru, K., Maeda, K., Harada, S. ,Mitsuya, H.

(2002) Identi cation of amino acid residues critical for LD78,

a variant of human macrophage in ammatory protein-1,

binding to CCR5 and inhibition of R5 human immunode -

ciency virus type 1 replication. Journal of Biological Chemis-

try, Vol. 277, No. 7, pp. 4649-55.

Murphy, P.M. (2001) Viral exploitation and subversion of the

immune system through chemokine mimicry. Nature immu-

nology, Vol. 2, No. 2, pp. 116-22.

Nakayama, K., Nakamura, H., Koga, M., Koibuchi, T., Fujii, T.,

Miura, T., Iwamoto, A. ,Kawana-Tachikawa, A. (2012) Imbal-

anced production of cytokines by T cells associates with the

activation/exhaustion status of memory T cells in chronic HIV

type 1 infection. AIDS research and human retroviruses, Vol.

28, No. 7, pp. 702-14.

Öztürk, O., Aksaç, A., Elsheikh, A., Özyer, T. ,Alhajj, R. (2013) A

consistency-based feature selection method allied with linear

SVMs for HIV-1 protease cleavage site prediction. PloS one,

Vol. 8, No. 8, p. e63145.

Page, J., Louw, M. ,Pakkiri, D. 2006, Working with HIV/Aids,

Juta and Company Ltd.

Pancholi, N.J., Gupta, S., Sapre, N. ,Sapre, N.S. (2014) Design

of novel leads: ligand based computational modeling stud-

ies on non-nucleoside reverse transcriptase inhibitors

(NNRTIs) of HIV-1. Molecular BioSystems, Vol. 10, No. 2,

pp. 313-25.

Park, B.J., Wannemuehler, K.A., Marston, B.J., Govender, N.,

Pappas, P.G. ,Chiller, T.M. (2009) Estimation of the current

42 MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS

Sweta, Usha and Sunil

global burden of cryptococcal meningitis among persons liv-

ing with HIV/AIDS. Aids, Vol. 23, No. 4, pp. 525-30.

Pavlasova, G., Borsky, M., Seda, V., Cerna, K., Osickova, J.,

Doubek, M., Mayer, J., Calogero, R., Trbusek, M. ,Pospisilova,

S. (2016) Ibrutinib inhibits CD20 upregulation on CLL B cells

mediated by the CXCR4/SDF-1 axis. Blood, Vol. 128, No. 12,

pp. 1609-13.

Pawlowski, A., Jansson, M., Sköld, M., Rottenberg, M.E. ,Käl-

lenius, G. (2012) Tuberculosis and HIV co-infection. PLoS Pat-

hog, Vol. 8, No. 2, p. e1002464.

Poongavanam, V., Kongsted, J. (2013) Virtual screening models

for prediction of HIV-1 RT associated RNase H inhibition. PloS

one, Vol. 8, No. 9, p. e73478.

Prosperi, M., De Luca, A. (2011) Computational models for pre-

diction of response to antiretroviral therapies. AIDS reviews,

Vol. 14, No. 2, pp. 145-53.

Prosperi, M.C., Altmann, A., Rosen-Zvi, M., Aharoni, E., Bor-

gulya, G., Bazso, F., Sönnerborg, A., Schülter, E., Struck, D.

,Ulivi, G. (2009) Investigation of expert rule bases, logistic

regression, and non-linear machine learning techniques for

predicting response to antiretroviral treatment. Antivir Ther,

Vol. 14, No. 3, pp. 433-42.

Riemenschneider, M., Heider, D. (2016a) Current Approaches in

Computational Drug Resistance Prediction in HIV. Current HIV

research, Vol. 14, No. 4, pp. 307-15.

Riemenschneider, M., Senge, R., Neumann, U., Hüllermeier,

E. ,Heider, D. (2016b) Exploiting HIV-1 protease and reverse

transcriptase cross-resistance information for improved drug

resistance prediction by means of multi-label classi cation.

BioData mining, Vol. 9, No. 1, p. 10.

Saini, V., Marchese, A. ,Majetschak, M. (2010) CXC chemokine

receptor 4 is a cell surface receptor for extracellular ubiquitin.

Journal of Biological Chemistry, Vol. 285, No. 20, pp. 15566-

76.

Samson, M., Labbe, O., Mollereau, C., Vassart, G. ,Parmentier,

M. (1996) Molecular cloning and functional expression of a

new human CC-chemokine receptor gene. Biochemistry, Vol.

35, No. 11, pp. 3362-7.

Sander, O., Sing, T., Sommer, I., Low, A.J., Cheung, P.K., Har-

rigan, P.R., Lengauer, T. ,Domingues, F.S. (2007) Structural

descriptors of gp120 V3 loop for the prediction of HIV-1 core-

ceptor usage. PLOS Comput Biol, Vol. 3, No. 3, p. e58.

Scherzer, R., Estrella, M., Yongmei, L., Deeks, S.G., Grunfeld, C.

,Shlipak, M.G. (2012) Association of tenofovir exposure with

kidney disease risk in HIV infection. AIDS (London, England),

Vol. 26, No. 7, p. 867.

Singh, O., Su, E.C.-Y. (2016) Prediction of HIV-1 protease

cleavage site using a combination of sequence, structural, and

physicochemical features. Bmc Bioinformatics, Vol. 17, No. 17,

p. 279.

Slimani, H., Charnaux, N., Mbemba, E., Saffar, L., Vassy, R.,

Vita, C. ,Gattegno, L. (2003) Interaction of RANTES with syn-

decan-1 and syndecan-4 expressed by human primary mac-

rophages. Biochimica et Biophysica Acta (BBA)-Biomem-

branes, Vol. 1617, No. 1, pp. 80-8.

Srisawat, A., Kijsirikul, B. (2008) Combining classi ers for

HIV-1 drug resistance prediction. Protein and peptide letters,

Vol. 15, No. 5, pp. 435-42.

Struyf, S., Menten, P., Lenaerts, J.P., Put, W., D’Haese, A., De

Clercq, E., Schols, D., Proost, P. ,Van Damme, J. (2001) Diverg-

ing binding capacities of natural LD78 isoforms of mac-

rophage in ammatory protein‐1 to the CC chemokine recep-

tors 1, 3 and 5 affect their anti‐HIV‐1 activity and chemotactic

potencies for neutrophils and eosinophils. European journal of

immunology, Vol. 31, No. 7, pp. 2170-8.

Tebas, P., Stein, D., Tang, W.W., Frank, I., Wang, S.Q., Lee, G.,

Spratt, S.K., Surosky, R.T., Giedlin, M.A. ,Nichol, G. (2014)

Gene editing of CCR5 in autologous CD4 T cells of persons

infected with HIV. New England Journal of Medicine, Vol. 370,

No. 10, pp. 901-10.

Wawer, M.J., Gray, R.H., Sewankambo, N.K., Serwadda, D., Li,

X., Laeyendecker, O., Kiwanuka, N., Kigozi, G., Kiddugavu, M.

,Lutalo, T. (2005) Rates of HIV-1 transmission per coital act, by

stage of HIV-1 infection, in Rakai, Uganda. Journal of Infec-

tious Diseases, Vol. 191, No. 9,

pp. 1403-9.

Wei, Y., Li, J., Chen, Z., Wang, F., Huang, W., Hong, Z. ,Lin, J.

(2015) Multistage virtual screening and identi cation of novel

HIV-1 protease inhibitors by integrating SVM, shape, pharma-

cophore and docking methods. European journal of medicinal

chemistry, Vol. 101, pp. 409-18.

Wiskerchen, M. ,Muesing, M.A. (1995) Human immunode -

ciency virus type 1 integrase: effects of mutations on viral

ability to integrate, direct viral gene expression from unin-

tegrated viral DNA templates, and sustain viral propaga-

tion in primary cells. Journal of virology, Vol. 69, No. 1,

pp. 376-86.

Xuan, S., Wu, Y., Chen, X., Liu, J. ,Yan, A. (2013) Prediction

of bioactivity of HIV-1 integrase ST inhibitors by multilinear

regression analysis and support vector machine. Bioorganic &

medicinal chemistry letters, Vol. 23, No. 6, pp. 1648-55.

Yan, A., Xuan, S. ,Hu, X. (2012) Classi cation of active and

weakly active ST inhibitors of HIV-1 integrase using a support

vector machine. Combinatorial chemistry & high throughput

screening, Vol. 15, No. 10, pp. 792-805.

Zazzi, M., Cozzi-Lepri, A. ,Prosperi, M.C. (2016) Computer-

Aided Optimization of Combined Anti-Retroviral Therapy for

HIV: New Drugs, New Drug Targets and Drug Resistance. Cur-

rent HIV research, Vol. 14, No. 2, pp. 101-9.

Zazzi, M., Incardona, F., Rosen-Zvi, M., Prosperi, M., Lengauer,

T., Altmann, A., Sonnerborg, A., Lavee, T., Schülter, E. ,Kai-

ser, R. (2012) Predicting response to antiretroviral treatment by

machine learning: the EuResist project. Intervirology, Vol. 55,

No. 2, pp. 123-7.

Zhen, A. ,Kitchen, S. (2013) Stem-cell-based gene therapy for

HIV infection. Viruses, Vol. 6, No. 1, pp. 1-12.

BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS MACHINE LEARNING APPROACHES TO STUDY HIV/AIDS INFECTION 43