Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
15
,
No.
1
,
Febr
uary
20
25
, pp.
1
132
~
1141
IS
S
N:
20
88
-
8708
, DO
I: 10
.11
591/ij
ece.v
15
i
1
.
pp
1132
-
1
141
1132
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om
Implem
enti
ng
c
l
oud
com
puti
ng
in
d
ru
g
d
is
covery
and
t
eleme
dicin
e for
q
ua
ntita
tive
s
t
ru
cture
-
a
ct
ivity
r
elati
ons
hip
a
nal
ysis
Pala
y
anoo
r
Se
eth
ap
athy
R
am
ap
r
aba
1
,
Be
l
lam
Ravindr
a B
ab
u
2
,
Na
ll
at
ha
mpi
R
ajam
an
i
Rej
in
Paul
3
,
V
ar
adan
Sh
ar
mi
la
4
,
Ve
nk
atach
ala
m
R
am
esh Ba
bu
5
,
R
am
an
R
amya
6
,
Subb
i
ah
Mur
ugan
7
1
Dep
artm
en
t of
E
l
ectrica
l
and
E
lect
r
o
n
ics Eng
in
eering
,
Pan
im
ala
r
Eng
in
eering
Co
lleg
e,
Ch
en
n
ai,
Ind
ia
2
Dep
artm
en
t of
Co
m
p
u
ter
Sci
en
ce a
n
d
E
n
g
in
eering
Departmen
t,
Ad
a
m
a S
cien
ce a
n
d
T
echn
o
lo
g
y
Univ
ersity
,
Ad
am
a,
Ethio
p
ia
3
Dep
artm
en
t of
Co
m
p
u
ter
Sci
en
ce a
n
d
E
n
g
in
eering
,
R.
M.
K.
Co
lleg
e of
E
n
g
in
eering
and
T
echn
o
lo
g
y
,
Ch
en
n
ai
,
Ind
ia
4
Dep
artm
en
t of
Co
m
p
u
ter
Sci
en
ce a
n
d
E
n
g
in
eering
,
R.
M.
D.
E
n
g
in
eering
Co
lleg
e,
Ch
en
n
ai,
Ind
ia
5
Dep
artm
en
t of
Co
m
p
u
t
er
Sci
en
ce a
n
d
E
n
g
in
eering
,
Dr.
MGR
Edu
catio
n
al and
Research Ins
tit
u
te,
Ch
en
n
ai,
Ind
ia
6
Dep
artm
en
t of
E
l
ectron
ics an
d
Co
m
m
u
n
icatio
n
E
n
g
in
eering
,
Ch
en
n
ai I
n
stitu
te of T
echn
o
lo
g
y
,
Ch
en
n
ai,
Ind
ia
7
Dep
artm
en
t of
Bi
o
m
ed
ical E
n
g
in
eer
in
g
,
Sav
eeth
a
Sch
o
o
l of E
n
g
in
eering
,
Sav
eeth
a I
n
stitu
te
o
f
Medical
and
T
e
ch
n
ical Sciences,
Sav
eeth
a Univ
ersit
y
,
Ch
en
n
ai,
Ind
ia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
M
a
y 13,
2024
Re
vised
A
ug 8, 20
24
Accepte
d
Se
p 3, 2
024
Thi
s
work
a
im
s
to
use
cut
ti
ng
-
edge
m
ac
hin
e
learni
ng
methods
to
im
prove
quant
itati
v
e
stru
ct
ure
-
ac
t
ivi
ty
relati
onship
(QS
AR)
ana
lysis
,
whi
ch
is
used
in
drug
dev
el
op
ment
and
telemedi
ci
ne
.
The
ma
jo
r
goa
l
is
to
ex
am
in
e
the
per
forma
n
ce
of
seve
ral
pr
edicti
v
e
mode
li
ng
app
r
oac
hes,
includin
g
ran
dom
fore
st
,
d
ee
p
le
a
rning
-
base
d
QS
AR
mode
ls,
an
d
s
upport
v
ecto
r
m
ac
h
ine
s
(SV
M)
.
It
inve
stiga
t
es
the
p
ote
ntial
of
f
eature
sel
ec
t
ion
te
chn
ique
s
deve
lop
ed
in
che
mo
inform
a
tics
for
enh
anci
ng
model
a
ccurac
y.
Th
e
innova
ti
v
e
aspe
ct
is
using
c
l
oud
computing
resourc
es
to
strengt
h
en
com
put
at
ion
al
s
kil
ls,
allowing
f
or
ma
n
agi
ng
mas
sive
am
oun
ts
o
f
chemical
inform
a
ti
on.
Th
is
strat
egy
pro
duce
s
accurate
and
gene
r
aliz
abl
e
QS
AR
mode
ls.
By
usi
ng
th
e
cl
oud's
sca
l
abi
l
it
y
and
c
onstant
availa
b
ility,
re
mote
hea
l
thc
ar
e
apps
have
a
workabl
e
answer.
The
goal
is
to
show
how
the
se
me
thods
m
ay
i
mprove
te
l
emed
ic
in
e
and
th
e
d
rug
developme
n
t
proc
ess.
Util
izing
cl
oud
com
puti
ng
equi
p
s
rese
arc
he
rs
wit
h
a
f
le
x
ibl
e
se
t
of
too
ls
for
pre
ci
se
and
time
ly
QS
AR
ana
lys
is,
spe
edi
ng
up
the
discov
ery
of
bio
ac
t
ive
che
m
ic
a
ls
for
th
era
peu
ti
c
use.
T
his
new
method
fit
s
wel
l
with
t
he
dyna
m
ic
nat
ure
of
ph
armac
eu
ti
c
al
study
and
has
the
pot
ent
i
al
to
tra
nsfo
rm
th
e
way
drugs a
re
develo
ped
and
de
li
ver
e
d
to
p
at
i
ent
s v
ia
te
l
em
ed
ic
in
e.
Ke
yw
or
d
s
:
Chem
oin
f
orma
ti
cs
Cl
oud
c
ompu
ti
ng
Drug d
isc
over
y
Qu
a
ntit
at
ive stru
ct
ur
e
-
act
ivit
y
Re
la
ti
on
sh
i
p
a
nalysis
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
BY
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
Pala
yano
or Se
et
hap
at
hy Ra
m
apr
a
ba
Dep
a
rtme
nt of
Ele
ct
rical
an
d
Ele
ct
ro
nics
E
nginee
rin
g
,
Pani
mala
r
E
ng
i
neeri
ng
C
ollege
Ba
ng
al
or
e
Tr
unk R
oa
d, Va
ra
dh
a
rajap
ur
a
m,
Poon
a
mall
ee,
Chen
nai, Ta
mi
l Nadu
6001
23,
I
ndia
Emai
l:
rama
prabatami
lse
lvan
@gmai
l.com
1.
INTROD
U
CTION
Cl
oud
c
omput
ing
has
e
vo
lved
a
s
a
ga
me
-
c
hangin
g
te
chnolo
gy
that
pro
vid
es
unmatc
he
d
com
pu
ta
ti
onal
powe
r
a
nd
stora
ge
ca
pac
it
y
to
var
i
ous
industries
,
includi
ng
te
le
medici
ne
a
nd
dru
g
dev
el
opment
.
Ex
plo
it
ing
cl
oud
c
omp
uting
resou
rces
has
s
ub
sta
ntial
ly
impro
ved
t
he
e
ff
i
cacy
an
d
acc
uracy
o
f
pr
e
dicti
ve
m
od
el
ing
i
n
the
co
ntext
of
qua
ntit
at
ive
struct
ur
e
-
act
ivit
y
relat
ion
s
hi
p
a
nalysi
s,
wh
ic
h
plays
a
vital
ro
le
in
ide
ntifyi
ng
ne
w
drug
s
[1]
.
T
his
is
bec
ause
qua
ntit
at
i
ve
struct
ur
e
-
ac
ti
vity
relat
io
nship
(
QSAR
)
an
al
ys
i
s
is
one
of
t
he
mo
st
imp
or
ta
nt
aspects
of
dr
ug
disco
very.
Re
searche
rs
ca
n
s
pee
d
up
the
anal
ys
is
of
m
assive
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Imple
men
ti
ng
c
loud c
omp
uting i
n
d
r
ug
d
isc
overy
and
t
el
emed
ic
ine
…
(
P
al
ayanoor
S
eet
hapathy
R
ama
praba
)
1133
dataset
s,
w
hic
h
ena
bles
the
disco
very
of
pros
pecti
ve
dru
g
can
did
at
es
w
it
h
impro
ved
a
ccur
ac
y.
T
his
i
s
made
po
s
sible
by
us
i
ng
t
he
scal
a
bili
ty
a
nd
fle
xib
il
it
y
of
cl
ou
d
plat
forms
.
Cl
oud
c
ompu
ti
ng
has
been
ad
opte
d
by
the
boom
i
ng f
ie
ld
of
tel
eme
dicin
e, which
seek
s
to brid
ge
the
gap
betwee
n he
al
thcare
pr
act
it
ion
e
rs
a
nd p
at
i
ents by
us
in
g
di
gital
pl
at
fo
r
ms.
T
his
al
lows
te
le
med
ic
ine
to
off
er
di
sta
nt
heal
thca
re
se
r
vices
s
m
oo
t
hly
[
2]
.
Th
e
dr
ug
dev
el
opment method h
as b
e
en
c
omplet
el
y
trans
forme
d
due
to
the
i
ncor
porati
on o
f
cl
oud
c
o
m
pu
ti
ng r
e
so
urces
into
QSAR
an
al
ys
is.
Eve
n
th
ough
t
he
y
ha
ve
their
us
es
,
t
ra
diti
on
al
Q
SA
R
a
ppro
ac
hes
oft
en
nee
d
help
t
o
meet
the
c
ompu
ta
ti
on
al
de
man
ds
imp
os
e
d
by
intric
at
e
bio
lo
gi
cal
interact
io
ns
a
nd
huge
c
hemical
li
brari
es
[
3]
.
Cl
oud
-
base
d
inf
rastr
uctu
res
make
it
easi
er
for
resea
rch
e
rs
to
w
ork
t
oget
he
r
on
pr
oject
s.
This
ma
kes
it
po
s
sible
f
or
res
earche
rs
in
dif
f
eren
t par
ts of
the
w
orl
d
t
o
w
ork
t
og
et
he
r
i
n
r
eal
-
ti
me,
w
hic
h
in
t
urn
e
ncou
rag
es
a li
vely
i
nterchang
e
of i
deas a
nd app
ro
ac
hes
[4]
.
Cl
oud
c
omp
uting
has
bee
n
a
dr
ivi
ng
el
e
ment
be
hind
the
rise
of
te
le
medici
ne
as
a
po
te
ntial
ly
disruptive
forc
e
in
the
healt
hc
are
b
us
iness
.
These
platf
or
ms
may
pro
vid
e
va
rio
us
te
le
medici
ne
se
rv
i
ces
to
patie
nts,
i
nclu
di
ng
rem
ote c
onsu
lt
at
ion
s a
nd re
al
-
ti
me monit
or
i
ng
of
vital
sign
s
[
5]
. Clo
ud
-
base
d
t
el
emedi
ci
ne
so
luti
ons i
ncr
e
ase the a
vaila
bi
li
ty o
f heal
thc
are service
s a
nd f
aci
li
ta
te
the
smooth
flo
w of medical
in
for
mati
on
amo
ng
healt
hc
are
prof
e
ssio
na
ls.
This
ai
ds
he
al
thcare
pro
vide
rs
in
diag
no
si
ng
diseas
es,
pr
edict
ing
tren
ds,
an
d
per
s
onal
iz
ing
t
reatme
nt
plans
,
ulti
mate
ly
re
vo
l
ution
iz
i
ng
how
healt
hca
r
e
is
delive
red
and
ex
per
ie
nc
ed
[6]
.
Althou
gh
us
in
g
cl
oud
c
omp
uting
in
Q
SAR
analy
sis
a
nd
te
le
medici
ne
has
le
d
t
o
sig
ni
ficant
le
aps
f
orward,
sever
al
obsta
cl
es
sti
ll
need
to
be
ove
rcome.
Gu
a
ra
nte
ei
ng
t
he
inter
opera
bili
ty
of
va
rio
us
cl
oud
-
base
d
s
yst
ems
and
a
pps
is
e
ssentia
l
for
e
nab
li
ng
s
moot
h
co
mm
un
ic
at
ion
a
nd
data
intercha
nge
ac
ro
ss
t
he
healt
hcar
e
ecos
ys
te
m
[
7]
.
Fo
r
t
he
f
or
es
eeable
f
utu
re
,
ongoin
g
resea
r
ch
en
dea
vors
are
co
nce
ntrati
ng
on
imp
rovi
ng
t
he
accurac
y
of
predict
ive
m
od
el
s,
opti
mizi
ng
cl
oud
c
ompu
ti
ng
al
gorith
ms
for
QSAR
a
na
lysis,
a
nd
ex
pl
or
i
ng
innov
at
ive
wa
ys
to
le
ve
ra
ge
cl
oud
-
base
d
te
chnolo
gies
f
or
te
le
medici
ne
app
li
cat
ions.
As
the
inte
gr
a
ti
on
of
cl
oud
co
mputi
ng
c
onti
nu
e
s
to
ev
olv
e
,
it
s
syner
gy
with
QSAR
anal
ys
is
an
d
te
le
medici
ne
ho
l
ds
the
pro
mise
of
sh
a
ping
a
f
uture
w
her
e
pe
rs
on
al
iz
ed
medi
ci
ne
a
nd
rem
ot
e
healt
hca
re
s
erv
ic
es
are
bo
t
h
a
possibil
it
y
an
d
a
global
reali
ty
[8]
.
Cl
ou
d
-
bas
ed
data
anal
ytics
hav
e
eme
r
ged
as
a
c
orn
ersto
ne
in
dru
g
dev
el
opmen
t
and
te
le
medici
ne.
These
a
nalytic
s p
r
ov
i
de
insi
ghts
i
nto
intric
at
e
bi
ologica
l
i
nteracti
on
s
a
nd pat
ie
nt
healt
hca
r
e
da
ta
that
ar
e
unmatc
hed
by
a
ny
ot
her
met
hod
[9]
.
Data
anal
ytics
play
an
e
ss
entia
l
par
t
in
unde
rstan
ding
pa
ti
ent
data
in
te
le
me
dicine.
T
his
e
na
bles
healt
hca
r
e
pr
act
it
ion
e
rs
to
rec
ognize
pa
tt
ern
s,
a
ntici
pa
te
il
lness
ou
t
breaks
,
and
de
vel
op
th
e
mo
st
e
ff
ect
iv
e
treat
ment
regi
mens
[10]
.
Cl
oud
c
omp
uting's
scal
abili
ty
and
cost
-
ef
fecti
ven
e
s
s
are
ke
y
a
dvan
ta
ges
in
dr
ug
dev
e
l
opment
a
nd
te
le
me
dicin
e
[11]
.
Pa
y
-
as
-
you
-
go
meth
ods
ma
y
be
use
d
by
researc
hers
a
nd
healt
hcar
e
orga
nizat
ion
s
,
a
ll
ow
in
g
the
m
t
o
pay
only
f
or
the
re
source
s
that
the
y
u
se.
This
strat
egy,
wh
ic
h
is
bo
t
h
c
os
t
-
eff
ect
ive
a
nd
de
mo
c
rati
zes
ac
cess
to
s
ophisti
cat
ed
co
mputi
ng
capa
bili
ti
es,
le
vels
the
pla
ying
fie
ld
f
or
resea
rchers
a
nd
healt
hc
are
insti
tuti
on
s,
re
gardless
of
ho
w
la
r
ge
th
ey
are
or
how
mu
c
h
money
the
y ha
ve
[12]
.
Cl
oud
ser
vice p
r
ovider
s
m
us
t
com
ply
w
it
h
hi
gh
r
eg
ulato
ry
r
eq
uireme
nts
s
uch
a
s
the
h
eal
th
ins
ur
a
nce
portabil
it
y
a
nd
acco
unta
bili
ty
act
in
t
he
Un
i
te
d
Stat
es
a
nd
the
ge
ne
ral
data
protect
io
n
re
gu
la
ti
on
i
n
E
uro
pe.
These
la
w
s
an
d
re
gula
ti
on
s
wer
e
e
nacted
t
o
pr
otect
the
pr
ivacy
a
nd
secu
rity
of
pa
ti
ent
healt
h
in
forma
ti
on
.
B
y
ens
ur
in
g
that
pa
ti
ent
data
is
mana
ged
i
n
a
man
ner
t
h
at
is
bo
t
h
et
hical
an
d
la
wful,
c
omp
li
ance
with
the
se
ru
le
s
helps
t
o
c
reate
co
nf
ide
nce
be
tween
healt
hca
re
pract
it
ion
e
r
s,
patie
nts
,
a
nd
prov
i
der
s
of
cl
oud
se
r
vices
[
13]
.
The
co
mb
i
nation
of
arti
fici
al
intel
li
gen
ce
(
AI)
a
nd
cl
oud
com
pu
ti
ng
is
on
t
he
c
usp
of
r
adical
ly
al
te
rin
g
t
he
traj
ect
ory
of
both
dru
g
resea
r
ch
a
nd
te
le
me
dicine
i
n
the
f
uture.
Pati
ents
may
get
rap
i
d
medical
a
dv
ic
e
,
bo
ok
appointme
nts,
an
d
recei
ve
presc
riptio
n
remin
ders
us
i
ng
AI
-
dri
ve
n
chat
bots
a
nd
vi
rtual
a
ssist
ants
i
n
te
le
medici
ne.
These
te
c
hnol
og
ie
s
are
pow
ered
by
cl
oud
-
base
d
mac
hin
e
-
le
arn
i
ng
al
gor
it
hm
s
[
14]
.
P
r
edict
iv
e
analyti
cs
that
AI
powe
rs
gi
ve
healt
hcar
e
pract
it
ion
ers
insi
gh
ts
int
o
their
patie
nts'
be
ha
vi
or
patte
r
ns
,
al
l
ow
i
ng
them
to
rea
ct
proacti
vel
y,
pr
even
ti
ng
il
lnes
ses
an
d
pro
mot
ing
healt
hy
li
ves.
As
A
I
c
onti
nu
e
s
to
ev
ol
ve
an
d
integrate
sea
ml
essly
with
cl
ou
d
c
ompu
ti
ng,
this
pr
om
ise
holds
the
pro
mise
of
creati
ng
a
healt
hcar
e
la
nd
scape
that
is
not
only
hi
gh
l
y
pe
rs
onal
iz
ed
bu
t
al
so
hi
gh
l
y
acce
ssi
ble
[15]
.
The
e
xtensi
ve
us
e
of
cl
oud
c
omp
uting
in
dru
g
de
vel
opment
a
nd
te
le
medici
ne
raise
s
seve
ral
imp
or
ta
nt
et
hical
and
so
ci
et
al
quest
io
ns
that
mu
st
be
thoro
ughl
y
in
ve
sti
gated
[
16]
.
Ov
e
rc
om
in
g
c
halle
ng
e
s
in
in
te
gr
at
in
g
cl
oud
com
pu
ti
ng
for
enh
a
nce
d
qu
a
ntit
at
ive
struct
ur
e
-
act
ivit
y
relat
ion
s
hip
a
na
lysis
in
dru
g
disco
very
a
nd
te
le
medici
ne
:
Data
analyti
cs
and
c
omp
uting
are
t
ran
s
f
ormin
g
te
le
medici
ne
a
nd
dru
g
resea
rch.
QSAR
an
al
ys
is
is
esse
nt
ia
l
for
unde
rs
ta
nd
in
g
this
i
ndus
tr
y's
c
ompou
nd
chemical
str
uc
tures
an
d
b
i
olo
gical
act
ivit
ie
s.
Dataset
c
omplexit
y
a
nd
processi
ng
ma
ke
tra
diti
onal
QSAR
analysis
harde
r
.
Creat
ive
s
olut
ion
s
are
re
qu
i
red
t
o
ha
ste
n
medicat
io
n
de
velo
pm
e
nt
an
d
enh
a
nce
te
le
m
edici
ne
as
data
volu
me
an
d
c
omp
le
xity
rise.
Cl
oud
c
ompu
ti
ng
ma
y
im
pro
ve
Q
SA
R
a
n
a
lysis.
T
he
cl
oud
le
ts
researc
hers
m
anag
e
e
norm
ous
a
mou
nts
of
data
an
d
do
so
phist
ic
at
ed
r
eal
-
ti
me
cal
cul
at
ion
s,
s
peed
i
ng
up
medicat
io
n
ca
ndidate
sel
ect
io
n
a
nd
imp
rovi
ng
te
le
medici
ne
proto
col.
Cl
oud
co
mputi
ng
for
QSAR
anal
ys
is
is
diff
ic
ult.
The
la
ck
of
sta
n
da
rd
iz
e
d
cl
oud
QSAR
analy
sis
too
ls
are
a
pro
blem.
F
ra
mew
orks
with
cl
oud
infr
a
struct
ur
es
pr
oduce
a
naly
ti
cal
ineff
ic
ie
nc
ie
s
an
d
c
on
si
ste
ncy.
Secu
rity
a
nd
pri
vac
y
of
se
ns
it
ive
m
edical
data
durin
g
cl
oud
-
based
a
naly
sis
are
ke
y
c
on
siderati
ons.
U
na
uthorize
d
acc
es
s,
data
br
eac
hes,
an
d
re
gu
l
at
ory
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
1
132
-
1
141
1134
com
pliance
m
us
t
be
ad
dr
e
ss
ed
to
e
sta
blish
cred
i
bili
ty
an
d
acce
ptabili
ty
for
ph
a
rmace
utica
l
and
te
le
medici
ne
cl
oud
-
base
d
Q
SA
R
a
nalysis.
QSAR
anal
ysi
s
dema
nd
s
str
ong
c
ompu
te
r
cl
us
te
rs
a
nd
rap
i
d
al
gorith
ms
f
or
com
plica
te
d
c
al
c
ulati
on
s.
Re
searche
rs
m
us
t
bala
nce
anal
yt
ic
al
pr
eci
sio
n
an
d
c
os
t
-
ef
fe
ct
ive
cl
ou
d
re
so
urce
util
iz
at
ion
.
Re
searche
rs
a
nd
healt
hcar
e
pr
ofessi
on
al
s
nee
d
to
le
ar
n
cl
oud
-
base
d
QSAR
analysis's
meri
ts
an
d
dow
ns
ides
,
w
hich
i
nh
i
bits
it
s
implement
at
ion
.
Her
e
,
cl
oud
c
ompu
t
ing
for
QSA
R
analysis
i
n
dru
g
dev
el
opment
a
nd
te
le
medici
ne
is
discusse
d.
The
project
i
ntends
to
s
moot
hly
inc
orp
or
at
e
cl
oud
-
based
QSAR
analysis
into
phar
maceuti
cal
and
te
le
medici
ne
e
nter
pr
ise
s
by
a
ddressin
g
method
ologica
l
issues,
e
nh
a
nc
ing
resou
rce
al
l
oc
a
ti
on
,
e
ns
uri
ng
rob
us
t
sec
uri
ty
,
a
nd
pro
mo
ti
ng
awa
ren
e
ss.
This
mig
ht
c
ha
ng
e
dru
g
disc
ov
e
r
y,
sp
ee
d up me
dicat
ion
devel
op
ment, a
nd im
prov
e
tel
emedici
ne,
i
ncr
easi
ng
world
wide
p
at
i
ent outco
mes.
Using
c
lo
ud
c
ompu
ti
ng
for
QSAR
in
d
r
ug
d
isc
over
y
a
nd
t
el
eme
dicin
e
:
Drug
de
vel
opment
a
nd
te
le
medici
ne
use
Q
SA
R
a
nal
ys
is
to
un
der
st
and
co
mp
li
cat
ed
c
hemical
-
bio
lo
gical
co
nne
ct
ion
s.
Th
e
in
du
st
ry
may
cha
nge
with
cl
oud
co
mputi
ng.
I
nteg
rati
ng
cl
ou
d
c
ompu
ti
ng
f
or
QSAR
a
nalysi
s
is
t
ough,
bu
t
thi
s
Re
search
offe
rs
no
vel
con
c
epts
an
d
prac
ti
cal
method
s
that
migh
t
trans
f
or
m
dru
g
dev
el
opment
and
te
le
medici
ne.
Stan
dardize
d
cl
oud
-
ba
sed
QSAR
anal
ys
is
meth
odol
og
ie
s
a
re
a
n
im
porta
nt
researc
h
con
t
rib
ution.
B
y
desig
ning
cl
oud
-
c
ompati
ble
processes
,
the
work
im
prov
e
s
analyti
cal
c
onsist
ency
,
ef
fici
ency,
and
reli
abili
ty
acro
s
s
dataset
s
.
These
st
rateg
ie
s
le
t
academic
s
and
pract
it
i
on
e
rs
us
e
cl
ou
d
co
mputi
ng
to
sp
eed
up
dru
g
re
sear
ch
an
d
im
pro
ve
te
le
medici
ne
.
This
e
nh
a
nce
s
the
pri
vac
y
a
nd
secu
rity
of
cl
oud
-
base
d
Q
SA
R
analysis.
R
obust
enc
r
yp
ti
on,
acce
ss
c
ontrol
s,
a
nd
re
gu
la
t
ory
co
mp
li
an
c
e
protect
me
di
cal
data.
T
he
study
addresses
t
hes
e
c
on
ce
rns
t
o
enh
a
nce
sta
ke
ho
l
der
c
onfide
nce
in
cl
oud
-
ba
sed
Q
SA
R
a
nalysis's
inte
gri
ty
a
nd
confide
ntial
it
y,
al
lowing
pha
rmace
utica
l
an
d
te
le
me
dicine
upta
ke.
T
his
study
reli
es
on
c
reati
ve
res
ource
op
ti
miza
ti
on.
Com
plex
QSAR
cal
culat
io
ns
an
d
c
os
t
-
e
ff
ect
ive
cl
ou
d
resour
ce
c
on
su
m
ptio
n
is
toug
h
to
reconcil
e.
B
y
opti
mizi
ng
a
lgorit
hm
s
an
d
co
mputi
ng
r
eso
ur
ces
,
Re
s
earch
yields
high
-
qual
it
y
r
esults
a
t
minimal
c
os
t.
This
e
ff
ect
ive
resou
rce
us
a
ge
ma
kes
cl
oud
-
base
d
QSAR
analysis
sust
ai
nab
le
,
scal
a
ble,
a
nd
afforda
ble for
academic
s a
nd
enter
pr
ise
s
o
f
a
ll
sizes. Cl
oud
-
base
d QSA
R
a
nalysis is
adv
oc
at
ed.
2.
LIT
ERATUR
E REVIE
W
Unde
rserve
d
gro
ups
ma
y
not
hav
e
acce
ss
to
the
essenti
al
te
chnolo
gy
an
d
i
nfrastr
uctu
re,
wh
ic
h
li
mit
s
their
abili
ty
to
par
ti
ci
pate
in
cl
oud
-
base
d
he
al
thcare
so
l
ution
s
.
T
his
pro
bl
em
is
caused
by
t
he
di
gital
div
ide
,
wh
ic
h
offers
a
so
ci
et
al
ba
rr
ie
r
.
T
o
bri
dge
t
his
ga
p,
deliberat
e
ef
forts
mu
st
be
made
to
e
na
ble
e
qu
it
able
a
ccess
to
cl
oud
-
based
te
le
medici
ne
serv
ic
es
[
17]
.
The
de
velo
pme
nt
of
te
le
me
di
ci
ne
an
d
drug
disco
ve
ry
will
li
kel
y
be
si
gn
ific
a
ntly
im
pacted
by
colla
borati
ve
r
esearch
pr
oject
s
ma
de
possibl
e
by
cl
oud
c
omp
uting
s
ys
te
ms.
T
his
sp
irit
of
c
oope
rati
on
helps
to
create
a
t
hr
i
vi
ng
resea
rc
h
c
om
m
unit
y
in
wh
ic
h
t
he
exc
hange
of
in
for
mati
on
helps
t
o
s
pee
d
up
t
he
disc
ov
e
r
y
of
new
sci
entifi
c
fin
dings
a
nd
im
pro
ves
the
cre
at
ion
of
break
thr
ough
treat
ments
[
18]
.
Op
e
n
-
acce
ss
databases
a
nd
cl
oud
-
base
d
pl
at
fo
r
ms
f
ur
t
her
de
moc
rati
ze
knowle
dge
by
gi
vin
g
sch
olars
acce
s
s
to
im
portant
res
ources
in
de
pende
nt
of
th
ei
r
physi
cal
lo
cat
ion
or
t
he
i
ns
ti
t
ution
s
the
y
are
a
ff
il
ia
te
d
with
[19]
.
Data
ce
nt
ers,
esse
ntial
t
o
the
op
e
rati
on
of
cl
oud
c
omp
uting,
are
r
esp
on
si
ble
for
a
la
rg
e
amo
un
t
of e
ne
r
gy cons
umpti
on
[20]
.
To
le
sse
n
the
ir
impact
on
the
en
vir
onme
nt,
s
uppliers
of
cl
oud
ser
vices
are
maki
ng
fi
nanci
al
inv
est
me
nts
in
ener
gy
-
sa
ving
te
ch
nolo
gy
and
in
vestigat
ing
no
vel
methods
s
uc
h
as
li
qu
id
c
ooli
ng.
T
he
healt
hcar
e
bus
iness
ma
y
le
ve
rag
e
the
pote
ntial
of
cl
oud
com
pu
ti
ng
res
pons
i
bly
by
a
doptin
g
met
hods
that
ben
e
fit
the
en
vi
ronme
nt.
This
will
li
nk
the
a
ims
of
te
c
hnol
ogy
grow
t
h
wi
th
en
vir
onment
al
con
se
rv
at
i
on
[21]
.
A
new
a
ge
of
in
novatio
n
a
nd
acce
ssi
bili
t
y
has
be
gun
with
t
he
us
e
of
cl
oud
c
ompu
ti
ng
in
qua
ntit
at
ive
structu
re
-
act
i
vity
co
nnect
ion
stud
ie
s
i
n
dr
ug
re
searc
h
a
nd
te
le
medici
ne
.
This
pro
mises
a
f
uture
in
wh
ic
h
per
s
onal
iz
ed
medici
ne
an
d
re
mo
te
healt
hcar
e
are
no
t
just
possibil
it
ie
s
bu
t
f
undame
ntal
ri
ghts
f
or
al
l
[2
2],
[
23]
.
The
us
e
of
c
loud
-
base
d
te
c
hnologies
in
dru
g
re
searc
h
a
nd
te
le
me
dicin
e
pr
e
sents
not
just
com
plex
le
gal
dif
ficult
ie
s
but
al
so
the
ne
ed
for
c
ompre
hensi
ve
regula
tor
y
f
rame
w
orks.
C
reati
ng
a
safe
env
i
r
onme
nt
f
or
cl
oud
-
base
d
healt
hca
re
ef
f
or
ts
a
nd
insti
ll
ing
tru
st
amo
ng
patie
nts,
healt
hcar
e
pro
vid
e
r
s,
an
d
researc
hers
al
i
ke
may
be
ac
com
plishe
d
vi
a
the
act
ive
purs
uit
of
inter
national
co
ope
rati
on
in
est
ab
li
sh
ing
un
i
form
la
ws
[24]
–
[26]
.
C
omp
uting
in
th
e
cl
oud
not
only
re
vo
l
utioni
zes
the
t
ec
hnologica
l
el
eme
nts
of
healt
hcar
e
but
al
so
al
te
rs
the
exp
e
rience
of
bein
g
a
patie
nt
.
Pati
ents
a
re
giv
e
n
m
ore
ag
encies
w
he
n
cl
oud
-
base
d
te
le
medi
ci
ne
te
ch
no
l
ogie
s
prov
i
de
th
em
easy
a
cces
s
to
me
dical
con
s
ultat
ion
s
,
i
nd
i
viduali
zed
healt
h
data,
a
nd
i
nteracti
v
e
to
ols
f
or
sel
f
-
monit
ori
ng.
Pati
ents
can
now
e
ng
a
ge
in
real
-
ti
m
e
co
ns
ultat
io
ns
wit
h
medical
s
pecia
li
sts,
seek
sec
ond
views
,
a
nd
act
ively
par
ti
ci
pate
in
t
he
tre
at
ment
plans
be
ing
de
vel
op
e
d
f
or
them,
tha
nks
to
sec
ure
cl
oud
c
onnecti
on
s
[
27]
,
[
28]
.
These
platfo
r
ms
al
lo
w
patie
nts
a
nd
heal
thcare
pr
act
it
ion
e
rs
t
o
remain
in
c
onsta
n
t
c
onta
ct
with
on
e
a
no
t
her,
re
su
lt
in
g
i
n
inc
rease
d
a
dhere
nce
t
o
trea
tment
reg
ime
ns
a
nd
impro
ved
ma
na
geme
nt
of
c
hro
nic
il
lnesses
.
N
ot
on
l
y
ca
n
cl
ou
d
-
base
d
te
le
medici
ne
cov
e
r
geog
raphical
bounda
ries,
but
it
al
so
de
epens
the
pat
ie
nt
-
pro
vid
e
r
connecti
on,
w
hich
helps
cr
eat
e
a
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Imple
men
ti
ng
c
loud c
omp
uting i
n
d
r
ug
d
isc
overy
and
t
el
emed
ic
ine
…
(
P
al
ayanoor
S
eet
hapathy
R
ama
praba
)
1135
colla
borati
ve
a
ppr
oach
to
t
he
delive
ry
of
he
al
thcare
[29]
.
T
h
e
d
e
t
a
i
l
e
d
s
t
u
d
y
f
o
l
l
o
w
s
t
h
e
p
r
o
p
o
s
e
d
s
y
s
t
e
m
,
r
e
s
u
l
t
s
a
n
d
d
i
s
c
u
s
s
i
o
n
,
c
o
n
c
l
u
s
i
o
n
.
3.
PROP
OSE
D SYSTE
M
3.1.
Cl
ou
d
co
mput
in
g
re
volutionizi
n
g
QS
AR
analy
sis i
n d
ru
g di
sco
ve
ry
a
nd
t
el
em
edic
ine
Cl
oud
c
ompu
t
ing
is
le
adin
g
dru
g
disco
ve
ry
an
d
te
le
me
dicine
tra
nsfo
r
mati
on
,
c
hang
ing
QSAR
analysis.
T
he
seaml
ess
i
ntegr
at
io
n
of
ne
w
c
ompu
ta
ti
on
al
meth
ods,
w
hich
pro
vi
de
ma
ny
benefit
s,
app
li
cat
io
ns
,
a
nd
a
dv
a
ntage
s
,
highli
gh
ts
it
s
imp
ort
ance
in
var
io
us
sec
tor
s.
Fi
gure
1
de
picts
c
oro
na
vir
us
disease
2019
(
COV
ID
-
19
)
re
la
te
d
A
I
-
base
d
me
dical
pictu
r
e
rec
ogniti
on
a
nd
a
nalysis.
H
ere,
i
nput
li
ke
X
-
ray
and
c
ompu
te
d
tomog
raph
y
(
CT
)
sca
n
data
set
s
are
cl
assi
f
ie
d
by
ap
ply
i
ng
a
dee
p
le
a
rn
i
ng
al
gorith
m
t
hro
ugh
image
process
i
ng and a
ugme
nt
at
ion
.
Figure
1. CO
V
ID
-
19
-
relat
ed deep
lear
ning
medical
ima
ge reco
gnit
ion
a
nd a
nalysis
sche
mati
c
3.2.
Dru
g
d
is
cov
ery
an
d
t
el
emedi
ci
ne
qu
an
ti
tative
s
tru
cture
-
ac
tivity
rel
at
io
nship
a
na
ly
sis
using
c
loud
c
omp
u
tin
g
Cl
oud
c
ompu
t
ing
is
disru
pting
me
dicat
ion
resea
rch
an
d
te
le
medici
ne.
I
t
has
tra
nsfo
r
med
QSAR
analysis
in
se
ve
ral
discipli
nes
,
c
hangin
g
how
aca
demics
a
nd
healt
hcar
e
prof
e
ssio
nals
a
ppr
oach
c
ompli
cat
ed
pro
blems
[
30]
.
This
tra
ns
it
io
n
re
volves
ar
ound
cl
oud
c
ompu
ti
ng.
It
has
r
evo
l
ution
iz
e
d
dru
g
de
vel
opm
ent
by
giv
in
g
resea
rc
her
s
acce
ss
to
massive
co
mputer
res
ources
and
c
utti
ng
-
ed
ge
meth
ods.
Fi
gure
2
s
hows
s
om
e
of
the
wa
ys
di
gital
te
chn
ol
og
ie
s
are
em
ployed
i
n
me
dicat
ion
de
velo
pm
e
nt.
T
his
fig
ure
co
nsi
sts
o
f
tw
o
di
vi
sion
s:
disco
very a
nd
dev
el
opment
.
Figure
2. A
n o
verview
of a
ppli
cat
ion
s of
digi
ta
l healt
h
te
ch
no
l
og
ie
s
in dr
ug
disco
very a
nd
dev
el
opme
nt
3.3.
Dr
ug
disc
ov
ery
and
teleme
dici
ne:
cl
oud c
omp
u
tin
g
f
or
Q
SAR
analy
sis wit
h r
andom
fores
t
In
QSAR
an
al
ys
is,
r
an
dom
f
or
est
(RF)
al
gorith
m
an
d
cl
oud
c
omp
utin
g
ha
ve
revolut
ion
iz
ed
dru
g
disco
very
a
nd
te
le
medici
ne.
RF
al
gorit
hm
a
nd
cl
ou
d
-
base
d
platf
orms
ha
ve
re
voluti
oni
zed
te
c
hniq
ues
i
n
these
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
1
132
-
1
141
1136
cru
ci
al
sect
or
s
by
imp
r
ov
i
ng
eff
ic
ie
nc
y,
acc
ur
ac
y,
an
d
c
re
at
ivit
y.
Cl
oud
com
pu
ti
ng
set
ti
ng
s
are
i
deal
f
or
RF
al
gorithms,
wh
ic
h
can
ma
na
ge
co
m
plica
te
d
i
nformat
ion a
nd d
e
velo
p reli
ab
le
p
re
dicti
on m
od
el
s
.
(
)
(
)
i
R
x
i
i
w
x
F
•
=
=
1
(1)
Eq
uation
(1)
presents
the
RF
f
or
QSAR
an
al
ys
is
(
)
,
w
hich
de
note
s
the
predict
io
n
f
or
a
gi
ve
n
input
,
is
the
num
ber
of
decis
ion
trees
i
n
the
forest,
repres
ents
the
re
gi
ons
def
i
ned
by
th
e
trees,
are
the
weig
hts
as
sign
e
d
to
eac
h
reg
i
on,
an
d
(
is
the
in
dicat
or
functi
on
deter
minin
g
if
fall
s
within
wh
e
n
harnesse
d
via
cl
ou
d
c
ompu
ti
ng,
t
his
al
go
rit
hm
ac
cel
erates
QSA
R
stud
ie
s,
e
na
bling
e
ff
ic
ie
nt
drug
disco
very
a
nd
te
le
medici
ne
a
dv
a
ncem
e
nts
t
hro
ugh
preci
se
act
ivit
y
pr
e
dic
ti
on
s
based
on
mo
le
cula
r
str
uc
tures.
QSAR
a
nalysi
s
in
dr
ug
disc
ov
e
r
y
a
nd
te
le
medici
ne
is
re
vo
l
ution
iz
e
d
by
cl
oud
co
mputi
ng,
a
s
s
ho
wn
in
Table
1.
RF
is
sta
ble
an
d
ac
cur
at
e,
dee
p
le
arn
i
ng
cat
ches
subtl
e
patte
r
ns,
s
upport
vector
mac
hin
es
(
SVM)
handle
high
-
di
mensi
on
al
data
w
el
l, a
nd ch
e
mo
in
f
or
mati
cs
-
base
d feat
ure
sel
ect
ion
im
prov
e
s inter
pret
abili
ty.
Table
1.
Cl
oud
-
p
owere
d
a
dv
a
nces i
n
QSAR
a
nalysis:
r
ev
ol
ution
iz
in
g
d
rug
d
isc
ov
e
ry an
d
t
el
emedici
ne
Ro
le
Ben
efit
Fu
n
ctio
n
Ap
p
licatio
n
Ad
v
an
tag
es
RF
for Q
SAR
an
al
y
sis
More ac
cu
rate
p
redictio
n
s
Decisio
n
tr
ee
en
sembles
in
crea
se
m
o
d
el
accuracy
.
Ph
arm
ac
eu
tical
d
isco
v
ery, tox
icity
forecast
Ro
b
u
st, complicated
data,
little ov
erf
ittin
g
Deep
learnin
g
-
b
as
ed
QSAR
m
o
d
el
Co
m
p
lex
pattern
recog
n
itio
n
deep
learnin
g
algo
rithm
s
Neu
ral
n
etwo
rks
le
arn
co
m
p
lex
data p
atte
rns
Desig
n
in
g
dru
g
s, v
irtual
screenin
g
,
p
redicti
n
g
co
m
p
o
u
n
d
pro
p
erti
es
Lear
n
s f
rom
big
datas
ets,
accurately
captu
re
s
n
o
n
lin
ear
co
n
n
ecti
o
n
s
SVM
for Q
SAR
an
aly
sis
Hig
h
-
d
im
en
sio
n
ally
eff
ectiv
e
Mapp
in
g
hig
h
-
d
im
en
sio
n
al data po
in
ts
find
s o
p
tim
u
m
h
y
p
erplan
e.
Predictin
g
bio
activ
ity
and
d
rug
-
target in
terac
tio
n
s
Go
o
d
gen
eraliza
tio
n
,
h
ig
h
-
d
im
en
sio
n
al data,
n
o
o
v
erf
ittin
g
Ch
em
o
in
form
atics
-
b
ased
f
eatu
re
sel
ec
tio
n
Fin
d
in
g
im
p
o
rtant
m
o
lecu
lar
ch
arac
teristi
cs
Selects i
m
p
o
rtant
ch
em
ical dat
a
ch
arac
teristi
cs
Ch
o
o
se fea
tu
res,
i
d
en
tify
m
o
lecu
lar
d
esc
ripto
rs
Mod
el interp
retabi
lity
,
d
im
en
sio
n
ality
r
ed
u
ctio
n
,
an
d
acc
u
racy i
m
p
r
o
v
ed
These
meth
ods
sp
ee
d
dru
g
di
sco
very
a
nd
e
na
ble
bi
oacti
vity
pr
e
dicti
ons
a
nd
m
olecular
de
sign
us
i
ng
cl
oud
platfo
r
ms.
Cl
oud
-
ba
sed
s
olu
ti
on
s
pro
vid
e
sea
mless
c
ollaborat
ion,
scal
a
bi
li
ty,
an
d
re
so
urce
op
ti
miza
ti
on,
making
the
m
essenti
al
for
phar
maceuti
cal
Re
searc
h
a
nd
te
le
medici
ne.
In
vitro
a
nd
s
il
ic
o
models
ena
ble
thera
peu
ti
c
ef
f
ect
iveness
a
nd
tox
ic
it
y
te
sti
ng
with
out
ani
mal
stud
ie
s.
In
vitro
m
odel
s
pro
vide
for novel
in
v
i
vo m
od
el
in
g. F
igure
3
e
xp
la
in
s drug sa
fety
a
nd ef
fecti
ve
nes
s test
.
The
fi
gure
s
um
ma
rizes
cu
rr
e
nt
nona
nim
al
m
od
el
opt
ion
s
for
sup
portin
g
ani
mal
awar
e
ness
requireme
nts,
i
nclu
ding
re
pla
ci
ng
,
lo
wer
in
g,
an
d
re
fining.
I
t
cov
e
rs
a
ppr
oa
ches
fo
r
re
pla
ci
ng
a
nimals
i
n
dr
ug
researc
h
to
ex
amine
ef
fects
and
reacti
ons
and
un
c
over
ge
netic
path
wa
ys
in
vo
l
ved
in
dr
ug
dev
el
opment
.
Chemic
al
abs
orptio
n,
distrib
ut
ion
,
meta
boli
sm,
e
xcr
et
io
n,
and
to
xicit
y
(
ADMET
),
enz
ym
e
as
say
,
s
pher
oid
culture
s
ys
te
m
s
,
QSAR
m
o
de
ls,
or
ganoid
models,
a
nd
s
ph
e
r
oid
c
ultu
r
e
syst
ems
im
pro
ve
dru
g
safe
ty
a
nd
eff
ic
acy
.
Figure
3. D
rug
saf
et
y an
d
e
ff
e
ct
iveness
te
st
i
den
ti
ficat
io
n,
protei
n,
bio
in
f
ormat
ic
s,
an
d mo
le
cular
dynami
cs
simulat
ion
s
sc
hemati
c
Dru
g
S
a
f
e
t
y
and
E
ff
icacy
AD
ME
T
A
ss
a
y
E
n
z
y
me
b
a
s
e
d
A
s
say
In
Si
li
c
o
S
c
r
e
e
n
i
n
g
S
p
h
e
r
o
i
d
C
u
ltur
e
S
y
s
te
ms
Org
a
n
o
id
Mo
d
e
l
Dru
g
P
r
o
t
e
c
t
ion
in
t
e
r
a
c
ti
o
n
QSAR
Mo
d
e
l
s
Dru
g
-
D
r
u
g
In
t
e
r
a
c
ti
o
n
s
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Imple
men
ti
ng
c
loud c
omp
uting i
n
d
r
ug
d
isc
overy
and
t
el
emed
ic
ine
…
(
P
al
ayanoor
S
eet
hapathy
R
ama
praba
)
1137
4.
RESU
LT
S
AND DI
SCUS
S
ION
4
.
1.
A
cl
oud
-
ba
sed
deep
le
arnin
g
-
ba
se
d
QSAR
m
od
el
fo
r
dru
g di
sco
very a
nd tel
e
medi
ci
ne
Cl
oud
c
ompu
ti
ng
an
d
d
ee
p
l
earn
i
ng
-
ba
sed
QSAR
m
od
el
s
ha
ve
revoluti
on
iz
e
d
dru
g
disco
very
a
nd
te
le
medici
ne.
This
s
yn
e
rgy
us
es
a
rtific
ia
l
neural
net
wor
ks
an
d
cl
oud
pl
at
fo
r
ms
to
re
de
fine
QSAR
a
nalysis.
This
te
ch
niqu
e
is
revoluti
oniz
ing
phar
ma
ceuti
cal
Re
search
a
nd
te
le
m
edici
ne
patie
nt
ca
re
by
st
udyi
ng
mo
le
cula
r
inte
r
act
ion
s a
nd
us
ing cl
oud
-
based com
puti
ng to
ol
s
[31
]
.
=
(
;
)
(2)
Eq
uation
(
2
)
s
hows
the
deep
le
arn
in
g
-
base
d
Q
SA
R
m
odel
,
wh
ic
h
re
pr
ese
nts
a
dru
g'
s
pr
e
dicte
d
qua
ntit
at
ive
act
ivit
y
,
deno
te
s
the
in
put
f
eat
ur
es
re
pres
enting
t
he
dru
g's
m
olecular
structu
re,
a
nd
re
presents
de
e
p
le
arn
in
g
m
od
e
l
pa
rameters
.
Cl
oud
c
omp
uting
power
s
t
he
m
odel
,
w
hich
proce
sses
huge
m
olecula
r
data
rem
otely
for
r
el
ia
ble
QS
AR
researc
h.
T
his
te
chn
iq
ue
imp
r
ov
e
s
pr
e
dicti
on
acc
ur
ac
y
f
or
dru
g
disco
ve
r
y
a
nd
te
le
medici
ne,
sp
ee
ds
up
c
he
mica
l
identific
at
ion
,
al
lows
ind
i
viduali
zed
patie
nt
treat
m
ents,
a
nd
tra
nsfo
rms
healt
hcar
e
.
4
.
2.
Usin
g
cl
oud c
omp
u
tin
g
f
or
Q
SAR
analy
sis i
n d
ru
g di
sco
very a
nd
t
el
emedi
ci
ne
using
SVMs
SVM
an
d
cl
oud
c
omp
uting
are
tra
nsfo
rm
ing
Q
SA
R
a
na
lysis.
The
st
r
eng
t
h
of
mac
hi
ne
le
ar
ning
al
gorithms
a
nd
cl
oud
c
omp
uting
is
tra
nsf
ormin
g
me
dicat
ion
resea
rc
h
and
te
le
me
dicine.
S
VM
a
nd
cl
oud
com
pu
ti
ng
ar
e
hel
ping
resea
rch
e
rs
i
nv
e
sti
ga
te
mo
le
c
ular
str
uctur
e
s,
ex
ped
it
e
medicat
ion
disc
over
y,
an
d
enh
a
nce
te
le
m
edici
ne
pa
ti
ent
care.
T
his
uni
qu
e
a
ppr
oac
h
us
es
S
V
M
al
gorith
ms
to
fin
d
and
predict
co
mp
le
x
data
patte
rns.
T
hese
al
gorit
hm
s
e
xamine
chemical
c
ompou
nd
s
a
n
d
predict
bio
l
og
ic
al
act
ivit
y
t
o
assist
medicat
io
n
de
velo
per
s
i
n
un
cov
e
rin
g
po
s
sibil
it
ie
s.
SV
M
models
delive
r
accurate
te
le
medici
ne
diag
no
sis
a
n
d
per
s
onal
iz
ed
t
reatment.
Hi
gh
-
dime
ns
io
nal
SVM
data
proces
sin
g
e
nhances
predict
ion
acc
ur
ac
y,
al
lowing
ta
rg
et
ed
d
rug d
isc
ov
e
ry an
d
te
le
medici
ne
t
he
rapy.
Tab
le
2 e
xp
la
in
s
cl
ou
d
-
powe
red en
ha
nc
ements
i
n QS
AR.
Figure
4
s
how
s
dru
g
disc
ov
e
ry
a
nd
te
le
me
di
ci
ne
QSAR
use
per
year.
Col
umns
s
how
ra
ndom
f
or
est
,
deep
le
a
rn
i
ng
-
base
d
QSAR
model,
S
VM,
and
c
hem
oinf
ormat
ic
s
-
based
featu
re
s
el
ect
ion
.
Data
shows
the
us
e
of
these
te
ch
niq
ue
s
f
r
om
2018
to
2020.
Ra
ndom
f
orest
a
nd
ch
em
oin
f
or
mati
cs
are
wide
ly
use
d,
s
ugge
sti
ng
their
popula
rity.
Table
2
. Cl
oud
-
p
owere
d
e
nh
a
nceme
nts in
QSAR
a
nalysis
Asp
ect
Ran
d
o
m
f
o
rest
for QS
AR
Deep
l
earnin
g
-
b
as
ed
QSAR
SVM
b
ased
QSAR
Ch
em
o
in
form
atics
-
b
ased
f
eatu
re
s
electio
n
Predictio
n
acc
u
rac
y
Hig
h
Very High
Hig
h
Mod
erate
Han
d
lin
g
complex
data
Yes
Yes
Yes
Yes
Interpretab
ility
Mod
erate
Low
Low
Hig
h
Co
m
p
u
tatio
n
al ef
fi
cien
cy
Fast
Mod
erate
Mod
erate
Fast
Featu
re
sele
ctio
n
c
ap
ab
ility
No
No
No
Yes
Figure
4. Com
par
at
ive
an
al
ysi
s o
f
QSAR tec
hn
i
qu
e
s in
drug
disco
very a
nd tel
eme
dicine
0
20
40
60
80
100
120
14
0
160
2018
2019
2020
R
a
n
d
o
m Fo
re
st
D
e
e
p
Learning
SVM
Chemoin
f
ormatics
Yea
rs
Feat
ures
Selec
tio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
1
132
-
1
141
1138
4
.
3
.
Qu
an
ti
t
at
i
ve
s
tr
uct
ur
e
-
a
c
tivity
r
el
ati
on
s
hip
a
na
l
ysi
s
in
d
rug
d
isc
overy
and
t
el
emedi
ci
ne
using
c
loud
-
b
as
ed
c
hemoin
fo
r
ma
t
ic
s
-
b
ase
d
f
ea
t
ure
s
el
ectio
n
The
f
us
i
on
of
chem
oinfo
rmat
ic
s
-
base
d
feat
ure
sel
ect
ion
te
c
hn
i
qu
e
s
with
c
loud
co
mputi
ng
so
l
utio
ns
has
r
ede
fine
d
t
he
la
ndsca
pe
of
Q
SA
R
a
nalys
is,
re
vo
l
ution
iz
ing
d
r
ug d
isc
over
y
a
nd
te
le
med
ic
ine p
racti
c
es.
By
harnessi
ng
the
powe
r
of
a
dvanced
c
hem
oinfo
rmati
cs
al
go
rithms
an
d
t
he
co
mputat
ion
al
scal
abili
ty
of
cl
oud
platfo
rms,
rese
arch
e
rs
a
nd
he
al
thcare
prof
es
sion
al
s
a
re
unr
aveli
ng
the
c
omplexit
ie
s
of
mo
le
cula
r
inter
act
ion
s
,
pav
i
ng
t
he
wa
y
f
or
gr
ound
breakin
g
ad
va
nc
ements
in
phar
maceuti
cal
r
es
earch
a
nd
pe
rs
on
al
iz
ed
patie
nt
care.
Chem
oin
f
orma
ti
cs
-
base
d
feat
ur
e
sel
ect
ion
s
erv
es
as
the
li
nc
hp
i
n
in
this
in
novative
a
ppr
oa
ch.
Re
sea
rc
he
rs
ca
n
us
e
co
mputat
ion
al
al
gorit
hms
to
i
den
ti
f
y
e
ssentia
l
m
olec
ular
de
script
ors
a
nd
st
ru
ct
ur
a
l
featu
res
cr
uc
ia
l
for
pr
e
dicti
ng
bi
olo
gical
act
ivit
ie
s.
I
n
dru
g
discov
e
r
y,
t
his
me
thod
strea
mli
ne
s
the
sel
ect
io
n
of
pote
ntial
dru
g
cand
i
dates,
al
l
ow
i
ng
resea
rc
her
s
t
o
f
oc
us
on
co
mpo
unds
with
the
hi
ghest
li
kelihood
of
s
uccess
.
RF
,
d
e
e
p
l
earn
in
g
-
b
ase
d
Q
SA
R
m
od
el
,
S
V
M
,
an
d
c
he
mo
i
nformat
ic
s
-
base
d
featur
e
sel
ect
i
on
a
re
c
omp
ared
in
Figure
5.
Eac
h
approac
h
is
te
ste
d
f
or
acc
ur
a
cy,
se
ns
it
ivit
y,
and
s
pecifici
ty.
RF
is
m
os
t
accurate
(
92%
),
the
n
chem
oinfo
rmat
ic
s
featu
re
sel
e
ct
ion
(
90%)
.
S
VM
(
88
%)
a
nd
dee
p
l
ear
ning
-
b
ase
d
Q
SA
R
(
89%)
a
re
m
uc
h
le
ss
accurate.
Met
hods h
ave
simil
ar s
e
ns
it
ivit
y
a
nd spec
ific
it
y.
Figure
5. Com
par
at
ive
a
nal
ysi
s o
f
QSAR
t
ec
hn
i
qu
e
s
5.
CONCL
US
I
O
N
This
exe
mp
li
fi
es
how
c
utti
ng
-
ed
ge
mac
hin
e
le
arn
in
g
met
hods
li
ke
RF
,
de
ep
le
ar
ning
-
ba
sed
Q
SA
R
models,
a
nd
s
upport
vecto
r
machi
nes
ma
y
re
voluti
oniz
e
QSAR
a
nalysis.
With
t
he
ad
diti
on
of
cl
oud
com
pu
ti
ng r
es
ources,
co
mput
at
ion
al
po
wer
i
s
furthe
r
am
plifie
d,
open
i
ng
the d
oor
to
ma
na
ging
la
r
ge
c
he
mica
l
dataset
s
an
d
imp
rovin
g
the
eff
ic
ie
nc
y
of
models.
T
his
novel
met
hod
not
only
has
the
po
te
ntial
to
ha
ste
n
the
dru
g
dev
el
opment
process
,
bu
t
it
al
s
o
s
ho
ws
promise
f
or
te
le
medici
ne
ap
plica
ti
on
s,
pro
vid
in
g
a
w
orka
ble
answ
er
to
t
he
pro
blem
of
he
al
thcare
deliv
ery
from
a
di
sta
nce.
T
o
ha
nd
le
inc
reasi
ngly
bigger
a
nd
m
or
e
com
plica
te
d
in
formati
on,
fu
t
ur
e
resea
rch
e
f
forts
sho
uld
re
fine
the
integ
r
at
ion
of
cl
oud
com
pu
ti
ng
res
ources
and
op
ti
mize
com
pu
ta
ti
onal
op
e
rati
ons.
Re
search
int
o
al
t
ern
at
ive
mac
hi
ne
le
a
rn
i
ng
al
gorithms
a
nd
un
i
qu
e
featur
e
sel
ect
ion
a
ppr
oach
es
in
ch
em
oin
f
or
mati
cs
mig
ht
l
ead
t
o
eve
n
be
tt
er
pre
dicti
on
model
acc
ur
ac
y
a
n
d
gen
e
rali
zabil
it
y.
In
ad
diti
on,
work
m
us
t
be
done
t
o
im
pro
ve
the
inter
pr
et
abili
ty
of
m
od
el
s
to
s
hed
li
ght
on
the
unde
rlying
c
he
mica
l
interac
ti
on
s.
T
his
f
ra
mew
ork
nee
ds
exte
ns
ive
val
idati
on
an
d
te
sti
ng
t
o
c
onfir
m
it
s
eff
ect
ive
ness
a
nd
de
penda
bili
ty
be
fore
bei
ng
use
d
i
n
real
-
world
cl
inica
l
sit
uations
.
C
omp
utati
on
al
c
he
mist
s,
ph
a
rmace
utica
l
resea
rc
her
s
,
a
n
d
healt
hcar
e
pro
vid
er
s
mu
st
w
ork
to
gethe
r
to
e
ns
ure
that
these
a
dv
a
nce
s
hel
p
patie
nts.
T
his
has
fa
r
-
reac
hing
co
ns
e
quence
s
for
im
prov
i
ng
healt
hc
are
outc
om
e
s
since
it
la
ys
the
path
for
a
new
way
of
thi
nk
i
ng
a
bout
dr
ug
researc
h
an
d
te
le
me
dicine.
I
n
the
f
uture
,
we
pro
vid
e
sec
ur
it
y
to
data
th
rou
gh
a cloud
sy
ste
m
.
70
75
80
85
90
95
Rand
o
m F
or
est
Dee
p Lear
ni
ng
-
B
a
s
ed QSAR
SV
M f
or
Q
SAR
Ana
l
ysis
Che
moinf
or
mat
i
c
s
Fe
a
t
ur
e Se
l
.
A
c
c
u
r
a
c
y
(
%)
Sen
sitiv
i
t
y (
%)
Sp
ec
i
fic
ity (
%)
Tec
h
n
iqu
es
A
c
c
u
rac
y,
Sensitivi
ty,
Speci
fic
ity(
%)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Imple
men
ti
ng
c
loud c
omp
uting i
n
d
r
ug
d
isc
overy
and
t
el
emed
ic
ine
…
(
P
al
ayanoor
S
eet
hapathy
R
ama
praba
)
1139
REFERE
NCE
S
[1]
N.
Kaso
ju
et
a
l.
,
“
Dig
ital
h
ealth
:
tr
en
d
s,
o
p
p
o
rtu
n
ities
an
d
ch
allen
g
es
in
m
ed
ical
d
ev
ices
,
p
h
arm
a
an
d
b
io
-
tech
n
o
lo
g
y
,
”
CS
I
Tra
n
sa
ctio
n
s o
n
I
CT
,
v
o
l.
1
1
,
n
o
.
1
,
p
p
.
1
1
–
3
0
,
2
0
2
3
,
d
o
i: 10
.10
0
7
/s4
0
0
1
2
-
023
-
0
0
3
8
0
-
3.
[2]
S
.
Su
g
an
d
h
a,
R
.
R
.
Ch
o
u
b
ey
,
R
.
K
.
Gu
p
ta,
an
d
S
.
B
.
Gu
p
ta
,
“
Ro
le
o
f
d
ig
ital
trans
form
atio
n
an
d
tech
n
o
l
o
g
y
ad
o
p
tio
n
in
t
h
e
eff
icien
cy
of
the p
h
arm
aceutic
al ind
u
stry
,
”
Eur
o
p
ea
n
Ch
emical
B
u
lletin
,
v
o
l.
1
2
,
p
p
.
6
8
6
2
–
6
8
7
4
,
2
0
2
3
.
[3]
S.
P.
S.
Rao
,
U
.
H.
Manju
n
ath
a,
S.
Miko
lajczak,
P.
G.
Ash
ig
b
ie,
an
d
T.
T.
Diag
an
a,
“
Drug
d
isco
v
ery
fo
r
p
arasitic
d
iseas
e
s:
p
o
were
d
b
y
tech
n
o
lo
g
y
,
en
ab
led
b
y
p
h
arm
acol
o
g
y
,
in
form
ed
b
y
clin
ical
scien
ce,
”
Tren
d
s
in
Par
a
sito
l
o
g
y
,
v
o
l.
3
9
,
n
o
.
4
,
p
p
.
2
6
0
–
2
7
1
,
2
0
2
3
,
d
o
i: 10
.10
1
6
/j.pt.
2
0
2
3
.0
1
.01
0
.
[4]
X.
Xu
et
a
l.
,
“
Me
ch
an
istic
stu
d
ies
o
n
alu
m
in
u
m
-
ca
taly
zed
ring
-
o
p
en
in
g
alternatin
g
co
p
o
ly
m
eri
zatio
n
o
f
m
al
eic
an
h
y
d
ride
with
ep
o
x
id
es:
lig
an
d
e
ff
ects
an
d
q
u
an
titativ
e
stru
ctu
r
e
-
activ
ity
rel
atio
n
sh
ip
m
o
d
el,
”
M
o
lecu
les
,
v
o
l.
2
8
,
n
o
.
2
1
,
2
0
2
3
,
d
o
i: 10
.3390
/m
o
lecu
les2
8
2
1
7
2
7
9
.
[5]
H.
J.
Hu
an
g
et
a
l.
,
“
Cu
rr
en
tly
u
sed
m
eth
o
d
s
to
ev
alu
ate
th
e
eff
ic
acy
o
f
th
erapeu
tic
d
rug
s
an
d
k
id
n
ey
safety,
”
Biomo
lecu
le
s
,
v
o
l.
1
3
,
n
o
.
1
1
,
2
0
2
3
,
d
o
i: 10
.3390
/b
i
o
m
1
3
1
1
1
5
8
1
.
[6]
S.
Thak
k
ar
et
a
l
.
,
“
Arti
ficial
in
telli
g
en
ce
an
d
real
-
w
o
rld
d
ata
for
d
rug
an
d
foo
d
safety
–
a
regu
lato
ry
sci
en
ce
p
ersp
ectiv
e,
”
Regu
la
to
ry To
xico
lo
g
y an
d
P
h
a
rma
c
o
lo
g
y
,
v
o
l.
1
4
0
,
May 2
0
2
3
,
d
o
i:
1
0
.1
0
1
6
/j.yrtp
h
.2023
.1
0
5
3
8
8
.
[7]
Q.
Ra
fiqu
e
et
a
l.
,
“
Rev
iewin
g
m
et
h
o
d
s
o
f
d
eep
lear
n
in
g
for
d
iag
n
o
si
n
g
COVID
-
1
9
,
it
s
v
ariants
an
d
sy
n
ergis
tic
m
ed
icin
e
co
m
b
in
atio
n
s,
”
Co
mp
u
ters
in Biolo
g
y an
d
Med
icin
e
,
v
o
l.
1
6
3
,
2
0
2
3
,
d
o
i: 1
0
.10
1
6
/j.compb
io
m
ed
.20
2
3
.10
7
1
9
1
.
[8]
S.
M
a
et
a
l.
,
“
Ma
ch
in
e
learnin
g
in
TCM
with
n
atu
ral
p
rod
u
cts
an
d
m
o
lecu
les:
cu
rr
en
t
sta
tu
s
an
d
futu
re
p
er
sp
ectiv
es,
”
Ch
in
ese
Med
icin
e (
Un
ited
Kin
g
d
o
m)
,
v
o
l.
1
8
,
n
o
.
1
,
2
0
2
3
,
d
o
i: 1
0
.11
8
6
/s
1
3
0
2
0
-
0
2
3
-
0
0
7
4
1
-
9.
[9]
K.
G.
Lock
wo
o
d
,
V.
Pit
ter,
P
.
R.
Ku
lk
arni
,
S.
A
.
Grah
am
,
L
.
A.
Au
ster
-
Gu
ss
m
an
,
an
d
O.
L.
H.
Bran
ch
,
“
Pr
ed
icto
rs
o
f
p
rog
ra
m
in
terest
in
a
d
ig
ital
h
ealth
p
ilo
t
stu
d
y
for
h
eart
h
ealth
,
”
PL
OS
Dig
ita
l
Hea
lth
,
v
o
l.
2
,
n
o
.
7
,
2
0
2
3
,
d
o
i: 10
.1371
/jo
u
rnal.p
d
ig
.00
0
0
3
0
3
.
[10
]
A.
K.
Bas
h
ir
et
a
l.
,
“
A
su
rvey
o
n
fed
erate
d
learnin
g
for
th
e
h
ealth
care
m
e
tav
erse:
co
n
cept
s,
ap
p
licatio
n
s,
ch
allen
g
es,
an
d
futu
re
d
irection
s,
”
a
rXiv
:
2
3
0
4
.0
0
5
2
4
,
p
p
.
1
–
1
5
,
2
0
2
3
.
[11
]
G.
Thah
n
iy
ath
et
a
l.
,
“
Clo
u
d
b
ased
p
r
ed
ictio
n
o
f
ep
ilep
tic
seizu
res
u
sin
g
re
al
-
tim
e
el
ectroen
ce
p
h
alo
g
ram
s
an
aly
sis
,
”
Inter
n
a
tio
n
a
l
Jo
u
rn
a
l
o
f
Electrica
l
a
n
d
Compu
ter
Eng
in
eerin
g
(I
JEC
E)
,
v
o
l.
1
4
,
n
o
.
5
,
p
p
.
6
0
4
7
–
6
0
5
6
,
Oct.
2
0
2
4
,
d
o
i: 10
.1159
1
/ijec
e.v1
4
i5
.pp
6
0
4
7
-
6
0
5
6
.
[12
]
Y.
Liu
et
a
l.
,
“
Inte
rpretable
ch
irality
-
awar
e
g
raph
n
eu
ra
l
n
etwo
rk
for
q
u
an
titativ
e
stru
ctu
re
a
ctiv
ity
relation
sh
ip
m
o
d
elin
g
in
d
rug
d
isco
v
ery,
”
Pro
ce
ed
in
g
s
o
f
th
e
3
7
th
AA
A
I
Co
n
feren
ce
o
n
Artificia
l
Intelli
g
en
ce,
A
AA
I
2
0
2
3
,
v
o
l.
3
7
,
p
p
.
1
4
3
5
6
–
1
4
3
6
4
,
2
0
2
3
,
d
o
i:
1
0
.16
0
9
/aaai.v3
7
i1
2
.26
6
7
9
.
[13
]
Y.
Matsu
zaka
an
d
Y.
Uesawa
,
“
Ens
em
b
le
learnin
g
,
d
e
ep
learnin
g
-
b
ased
an
d
m
o
lecu
lar
d
escript
o
r
-
b
ased
q
u
an
titativ
e
stru
ctu
re
–
activ
ity
r
elatio
n
sh
ip
s,
”
Mo
lecu
les
,
v
o
l.
2
8
,
n
o
.
5
,
2
0
2
3
,
d
o
i: 10
.3390
/m
o
lecu
les2
8
0
5
2
4
1
0
.
[14
]
X.
Hu
o
,
J.
Xu
,
M
.
Xu
,
an
d
H.
Ch
en
,
“
An
i
m
p
rov
ed
3
D
q
u
an
titativ
e
stru
ctu
re
-
activ
ity
relat
io
n
sh
ip
s
(QSAR)
o
f
m
o
le
cu
les
with
CNN
-
b
ased
p
art
ial
least
sq
u
ares
m
o
d
el,
”
Ar
tificia
l
Intellig
en
ce
in
th
e
Life
S
cien
ces
,
v
o
l.
3
,
2
0
2
3
,
d
o
i: 10
.1016
/j.ailsci.20
2
3
.10
0
0
6
5
.
[15
]
S.
J.
Belfi
eld
,
M.
T.
D.
Cro
n
in
,
S.
J.
Eno
ch
,
an
d
J.
W
.
Firm
a
n
,
“
Gu
id
an
ce
for
g
o
o
d
p
ractice
in
th
e
ap
p
licatio
n
o
f
m
achi
n
e
lea
rnin
g
in
d
ev
elo
p
m
en
t
o
f
to
x
ico
lo
g
ical
q
u
an
titativ
e
stru
ctu
re
-
activ
ity
relation
sh
ip
s
(QSARs)
,
”
PL
o
S
ONE
,
v
o
l.
1
8
,
n
o
.
5
May,
2
0
2
3
,
d
o
i: 10
.1371
/jo
u
rnal.p
o
n
e.02
8
2
9
2
4
.
[16
]
S.
H.
Park,
H
.
G.
Lee
,
X
.
Liu,
S.
K.
Lee,
an
d
Y
.
T.
Ch
an
g
,
“
Qu
an
titativ
e
stru
ctu
re
-
activ
ity
r
elatio
n
sh
ip
o
f
fluo
rescent
p
rob
es
an
d
th
eir
in
trace
llu
la
r
l
o
calization
s,
”
Ch
emo
sen
so
rs
,
v
o
l.
1
1
,
n
o
.
5
,
2
0
2
3
,
d
o
i: 10
.33
9
0
/ch
em
o
senso
rs1
1
0
5
0
3
1
0
.
[17
]
P.
Is
wan
to
,
I
.
M.
Firdau
s,
A
.
F
.
D
af
au
lh
aq
,
A.
G.
Ra
m
ad
h
an
i,
M.
P.
S
ap
u
tri,
an
d
H
.
Eko
wati,
“
Qu
an
titativ
e
stru
ctu
re
-
a
ctiv
ity
relation
sh
ip
o
f
3
-
t
h
io
cy
an
ate
-
1H
-
in
d
o
les
d
erived
co
m
p
o
u
n
d
s
as
an
tileu
k
em
ia
b
y
AM1,
P
M
3
,
an
d
RM1
m
e
th
o
d
s,
”
Ju
rn
a
l
Kimia
S
a
in
s d
a
n
A
p
lika
si
,
v
o
l.
2
6
,
n
o
.
3
,
p
p
.
10
9
–
11
7
,
2
0
2
3
,
d
o
i: 10
.14
7
1
0
/jk
sa.
2
6
.3.1
0
9
-
1
1
7
.
[18
]
T.
R.
No
v
ian
d
y
et
a
l.
,
“
Ens
em
b
le
m
a
ch
in
e
learnin
g
ap
p
roach
for
q
u
an
titativ
e
stru
ctu
re
activ
i
ty
relation
sh
ip
b
as
ed
d
rug
d
iscovery:
a r
ev
iew
,
”
Info
litik
a
Jou
rn
a
l of Da
ta
S
cien
ce
,
v
o
l.
1
,
n
o
.
1, pp
.
3
2
–
4
1
,
2
0
2
3
,
d
o
i: 10
.60
0
8
4
/ij
d
s.v
1
i1
.91
.
[19
]
D.
S
.
Megaw
ati,
J.
Eko
wati,
an
d
S.
Siswan
d
o
n
o
,
“
Qu
an
titativ
e
str
u
ctu
re
-
activ
ity
r
el
atio
n
sh
ip
(QSAR)
o
f
N
-
b
en
zo
y
l
-
N’
-
n
ap
h
ty
lth
i
o
u
rea
d
erivativ
e
co
m
p
o
u
n
d
s
b
y
in
silico
as
an
tican
cer
th
rou
g
h
in
h
ib
itio
n
o
f
VEG
FR2
rece
p
to
r
s,
”
in
Th
e
Inter
n
a
tio
n
a
l
Co
n
feren
ce on
Green Tech
n
o
lo
g
y an
d
E
n
erg
y E
n
g
in
eerin
g
,
2
0
2
3
,
p
p
.
1
3
7
–
1
4
8
,
d
o
i: 10
.2
9
9
1
/
9
7
8
-
94
-
6
4
6
3
-
1
4
8
-
7_15.
[20
]
R.
T
.
Pu
sp
arini,
A.
A
.
Krisn
ad
h
i,
an
d
Firday
an
i,
“
Math:
a
d
eep
lea
rnin
g
ap
p
roach
in
QSAR
for
estro
g
en
rece
p
to
r
alp
h
a
in
h
ib
ito
rs,
”
Mo
lecu
les
,
v
o
l.
2
8
,
n
o
.
1
5
,
2
0
2
3
,
d
o
i: 10
.3
3
9
0
/m
o
lecu
les2
8
1
5
5
8
4
3
.
[21
]
R.
Had
an
u
,
S.
A
b
d
Sam
ad
,
an
d
M.
Fath
Azzajad,
“
Qu
an
titativ
e
rel
atio
n
sh
ip
s
an
aly
sis
o
f
stru
ctu
re
an
d
activ
ity
o
f
asam
-
5
-
aryled
en
e
-
N,N
’
-
d
i
m
eth
y
lb
arbitu
ric
d
erivativ
es
as
an
u
ric
acid
d
rug
,
”
S
CIRE
A
Jo
u
rn
a
l
o
f
M
ed
icin
e
,
v
o
l.
7
,
n
o
.
2
,
p
p
.
1
–
1
6
,
2
0
2
3
,
d
o
i: 10
.5464
7
/
p
m
3
1
0
2
0
5
.
[22
]
S.
S.
Du
ay
,
R
.
C.
Y.
Yap
,
A.
L.
G
aitan
o
,
J.
A
.
A
.
S
an
to
s,
an
d
S.
J.
Y
.
Ma
calin
o
,
“
Ro
le
s
o
f
v
irtual
sc
reen
in
g
an
d
m
o
lecu
lar
d
y
n
am
ics
simulati
o
n
s
in
d
isco
v
ering
an
d
u
n
d
erstand
in
g
an
tim
alari
al
d
rug
s,
”
Inter
n
a
tio
n
a
l
J
o
u
rn
a
l
o
f
Mo
lecu
la
r
S
cien
ces
,
v
o
l.
2
4
,
n
o
.
1
1
,
2
0
2
3
,
d
o
i:
1
0
.33
9
0
/ijms24
1
1
9
2
8
9
.
[23
]
C.
Lesta
ri,
E
.
Da
r
win
,
D.
P
.
Pu
tra
,
N.
Su
h
arti,
an
d
B.
A.
Gan
i,
“
The
ro
le
o
f
p
lan
t
ex
tract
s
in
th
e
repair
o
f
r
attu
s
n
o
rveg
icu
s
m
an
d
ib
u
lar
alv
eo
lar
b
o
n
e
in
a
p
eri
o
d
o
n
titis
m
o
d
el,
”
Ras
a
ya
n
Jo
u
rnal
o
f
Ch
emistr
y
,
v
o
l.
1
6
,
n
o
.
3
,
p
p
.
1
3
4
2
–
1
3
5
0
,
2
0
2
3
,
d
o
i: 10
.3178
8
/RJ
C
.20
2
3
.1638
1
5
9
.
[24
]
T.
R.
No
v
ian
d
y
,
A.
Maulan
a,
T.
B
in
Emran
,
G
.
M.
Idro
es,
an
d
R.
Idr
o
es,
“
QSAR
class
ification
o
f
b
eta
-
secretase
1
in
h
ib
ito
r
activ
ity
in
alzh
eim
er
’s
d
iseas
e
u
sin
g
en
semble
m
achi
n
e
learnin
g
alg
o
rithms,
”
He
ca
Jo
u
rn
a
l
o
f
App
lied
S
cien
ces
,
v
o
l.
1
,
n
o
.
1
,
p
p
.
1
–
7
,
2
0
2
3
,
d
o
i:
10
.60
0
8
4
/h
jas.v
1
i
1
.12
.
[25
]
C.
Dh
an
d
e,
D.
Mi
stry
,
A
.
Karthic
,
R
.
Sin
g
h
,
an
d
S.
Ba
rage,
“
Co
m
p
u
tatio
n
al
ap
p
roach
es
to
i
d
en
tify
n
o
v
el
in
h
ib
ito
rs
for
th
e
d
rug
‐
resistan
t
Myco
b
acter
iu
m
tu
b
erculo
sis
Dp
rE
1
en
zy
m
e
,
”
Ind
o
n
esian
Jo
u
rn
a
l
o
f
B
io
tech
n
o
lo
g
y
,
v
o
l.
2
8
,
n
o
.
3
,
p
p
.
1
8
0
–
1
9
0
,
2
0
2
3
,
d
o
i: 10
.2214
6
/ijb
i
o
tech
.80
1
4
5
.
[26
]
A.
P.
To
ro
p
o
v
a,
A.
A.
Toro
p
o
v
,
A.
Ro
n
cagl
io
n
i,
an
d
E.
Ben
fenati,
“
Bin
d
in
g
o
rganop
h
o
sphate
p
esticid
es
to
acetylch
o
lin
esterase:
risk
ass
ess
m
en
t
u
sin
g
th
e
Mon
te
Carl
o
m
eth
o
d
,
”
To
xico
lo
g
ica
l
a
n
d
Envir
o
n
men
ta
l
Ch
emistr
y
,
v
o
l.
1
0
5
,
n
o
.
1
–
7
,
p
p
.
1
9
–
2
7
,
2
0
2
3
,
d
o
i: 10
.1080
/0
2
7
7
2
2
4
8
.2
0
2
3
.2
1
8
1
3
4
8
.
[27
]
T.
R.
Sa
ravan
an
,
A.
R.
Rath
in
am
,
J.
Lenin
,
A.
Ko
m
ath
i,
B.
Bh
arathi,
an
d
S.
Muru
g
an
,
“
Rev
o
lu
tio
n
izin
g
clo
u
d
co
m
p
u
tin
g
:
ev
alu
atin
g
th
e
in
fluen
ce
o
f
b
lo
ck
ch
ain
an
d
co
n
sen
su
s
alg
o
rithms,
”
in
2
0
2
3
3
rd
Inter
n
a
tio
n
a
l
Co
n
feren
ce
o
n
S
ma
rt
Gen
era
tio
n
Co
mp
u
tin
g
,
Co
mm
u
n
ica
tio
n
a
n
d
Net
wo
rkin
g
,
S
ma
rt Gen
co
n
,
2
0
2
3
,
d
o
i: 10
.11
0
9
/SMARTGE
NCON6
0
7
5
5
.2023
.10
4
4
2
0
0
8
.
[28
]
M.
J.
Ku
m
a
r,
S.
Mish
ra,
E.
G
.
Red
d
y
,
M.
Raj
m
o
h
an
,
S.
Muru
g
an
,
an
d
N.
A.
Vig
n
esh
,
“
Bay
esian
d
ecisio
n
m
o
d
el
b
ased
relia
b
le
rou
te
for
m
atio
n
in
in
ternet
o
f
th
in
g
s,
”
Ind
o
n
esia
n
Jo
u
r
n
a
l
o
f
Electrica
l
Eng
in
eerin
g
a
n
d
Co
mp
u
ter
S
cien
ce
(I
JEE
CS
)
,
v
o
l.
3
4
,
n
o
.
3
,
p
p
.
1
6
6
5
–
1
6
7
3
,
2
0
2
4
,
d
o
i: 1
0
.1
1
5
9
1
/ijeecs.v
3
4
.i3.
p
p
1
6
6
5
-
1
6
7
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
omp E
ng,
V
ol.
15
, No
.
1
,
Febr
uary
20
25
:
1
132
-
1
141
1140
[29
]
A.
R
.
Rath
in
a
m
,
B.
S.
V
ath
an
i,
A.
Ko
m
ath
i,
J.
Lenin
,
B.
Bh
arathi,
an
d
S.
M.
Urug
an
,
“
A
d
v
an
ces
an
d
p
redictio
n
s
in
p
redictiv
e
au
to
-
scalin
g
an
d
m
ain
ten
an
ce
a
lg
o
rithms
for
clo
u
d
co
m
p
u
tin
g
,
”
in
2
nd
Inter
n
a
tio
n
a
l
Confer
en
ce
o
n
Auto
ma
tio
n
,
Co
mp
u
tin
g
a
n
d
Renew
a
b
le Sys
tem
s, I
CAC
RS 2
0
2
3
-
Pro
ceedin
g
s
,
2
0
2
3
,
p
p
.
3
9
5
–
4
0
0
,
d
o
i:
10
.11
0
9
/ICACR
S
5
8
5
7
9
.20
2
3
.10
4
0
4
1
8
6
.
[30
]
M.
D.
A.
Hasan
,
K.
Balas
u
b
ad
ra,
G
.
Vad
iv
el,
N.
A
run
fr
ed
,
M
.
V
Ish
wa
r
y
a,
an
d
S.
Muru
g
an
,
“
IoT
-
d
riven
im
a
g
e
recog
n
itio
n
for
m
icrop
lastic
an
aly
sis
in
wat
er
sy
stems
u
sin
g
co
n
v
o
lu
t
io
n
al
n
eu
ral
n
etwo
rks
,
”
in
2
0
2
4
2
nd
Inter
n
a
tio
n
a
l
Co
n
fe
ren
ce
o
n
Co
mp
u
ter
,
Co
mmu
n
ica
tio
n
a
n
d
Co
n
tro
l
,
2
0
2
4
,
d
o
i: 10
.11
0
9
/IC4
5
7
4
3
4
.2024
.1048
6
4
9
0
.
[31
]
M.
A
m
ru
et
a
l.
,
“
Netwo
rk
in
trus
io
n
d
etectio
n
sy
ste
m
b
y
ap
p
ly
in
g
en
se
m
b
le
m
o
d
el
for
s
m
art
h
o
m
e
,
”
Int
er
n
a
tio
n
a
l
Jo
u
rn
a
l
o
f
Electrica
l an
d
Co
mp
u
ter Eng
in
eering
(I
JEC
E)
,
v
o
l.
1
4
,
no
.
3
,
p
p
.
3
4
8
5
–
3
4
9
4
,
2
0
2
4
,
d
o
i: 10
.
1
1
5
9
1
/ijece.v1
4
i3
.
p
p
3
4
8
5
-
3
4
9
4
.
BIOGR
AP
HI
ES OF
A
UTH
ORS
Pal
ayanoor
Se
e
thap
athy
Rama
praba
is
P
rof
essor
and
H
ea
d
of
the
Dep
art
m
e
nt
of
Elec
tr
ical
an
d
Elec
tron
ic
s
E
ngine
er
i
ng
has
t
he
t
ea
ch
ing
e
xp
eri
en
ce
of
22
y
e
ars
with
12
yea
rs
of
rese
arch
expe
r
ie
n
ce.
H
er
r
ese
arc
h
in
te
r
est
is
image
pro
ce
ss
ing,
me
d
ic
a
l
imaging,
em
bedd
ed,
inter
net
of
thi
ngs
,
an
d
ren
ew
abl
e
en
e
rgy.
She
has
gr
a
duat
ed
from
SR
M
ea
sw
ari
engi
ne
eri
ng
co
llege,
Madr
as
un
ive
rsity
Chenn
ai
in
el
e
ct
r
ic
a
l
an
d
e
lectr
oni
cs
en
gine
er
ing
(2000).
She
h
as
obta
in
ed
h
er
m
a
sters
in
Sathy
ab
am
a
Univer
si
ty
Chenna
i
(2006)
and
Ph.D.
d
egr
ee
a
lso
in
Sathya
b
am
a
Univ
ersit
y
Ch
enna
i
.
She
has
s
t
art
ed
her
edu
cation
carrie
r
in
the
yea
r
2000
as
a
l
ecture
r
.
She
h
a
s
ac
co
mpl
ished
an
ground
br
eaking
r
ese
arc
h
o
n
the
e
arl
y
“
Diagnosis
of
C
erv
ical
Cancer
u
sing
Colposcopic
Im
age
s
”
.
Thi
s
a
stute
r
ese
arc
h
h
as
brought
her
the
pride
of
owning
d
o
ct
or
at
e
in
eng
ineeri
ng
and
bio
te
ch
nology
.
She
h
as
jubilantl
y
par
ticipated
in
nume
rous
n
at
io
nal
and
i
n
te
rn
a
ti
onal
conf
er
ences
condu
cted
b
y
rep
u
te
d
orga
nizati
ons.
S
he
has
publi
she
d
seve
ral
p
ape
rs
i
n
i
nt
ern
ational
(18)
and
n
ation
al
(14)
and
al
so
pre
sent
ed
a
pape
r
in
i
nt
ern
a
ti
onal
conf
e
ren
c
e
and
n
at
ion
al
c
onfe
ren
c
es
(20).
She
is
the
me
mb
er
of
IE
E
E,
l
ife
me
mb
er
of
ISTE,
I
ET
E
a
nd
IEI
.
She
h
as
rec
e
ive
d
ou
tsta
n
ding
women
and
young
women
scie
n
t
ist
awa
rds.
She
ca
n
b
e
cont
a
ct
ed
at
e
ma
i
l:
ram
apr
aba
t
amils
el
van@gm
ai
l
.
co
m.
Bel
lam
Ravind
ra
Bab
u
obta
in
ed
B.
T
ech
.
from
Siddag
anga
Insti
tut
e
o
f
Te
chno
logy
(SIT),
Bang
al
or
e,
M.T
ec
h
.
and
Ph.
D.
fro
m
the
Jaw
aha
rl
al
Nehru
T
ec
hnolog
ic
a
l
Univer
sity,
Hy
der
aba
d
,
India
.
Wor
k
ed
in
s
ofware
industry
for
SIEMENS
,
Compu
te
r
As
socia
te
s,
YA
HO
O!
And
SA
P
L
abs
in
var
ious
rol
es
for
15
y
e
ars
.
Curre
nt
ly
h
e
h
as
b
ee
n
working
as
an
a
ss
oci
at
e
profe
ss
or
in
the
Depa
r
t
me
nt
of
Co
mputer
Sci
ence
and
Engi
ne
eri
ng
Depa
rtment,
Ad
am
a
Scie
n
ce
an
d
Technol
ogy
U
nive
rsity
(AS
TU
),
Adam
a,
Et
h
io
pia
.
He
has
tot
ally
23
y
ea
rs
of
exp
eri
en
ce
i
n
the
sof
twar
e
i
ndustry
and
as
an
a
ca
d
em
i
ci
an
.
Published
var
ious
articles
in
rep
ut
ed
journ
al
s
and
v
ari
ous
abstra
c
ts
were
pre
sente
d
at
in
te
rna
ti
ona
l
conf
ere
n
ce
s.
Hi
s
are
as
of
inter
est
ar
e
m
ac
hin
e
le
a
rning,
optimizati
on
techn
i
ques,
SO
A,
virt
ualisati
o
n,
cloud
com
put
ing,
distri
but
ed
sys
t
em
s,
softwar
e
d
esign
and
arc
h
itect
ur
e
and
com
put
er
ne
twor
ks
.
He can
be
co
nta
c
te
d
at e
m
ail:
Ravi
ndr
a.
b
abu
@astu.
edu
.
et
.
Nallathampi
Rajaman
i
Reji
n
Paul
is
an
ass
oci
a
te
prof
essor
in
the
Depa
r
tm
e
nt
of
CS
E,
R.
M.K
Coll
ege
o
f
En
gine
er
ing
and
Te
chno
logy,
Ch
enna
i
.
Compl
eted
UG
–
BE
c
omputer
s
c
ie
nc
e
and
e
ngineeri
ng
from
Anna
Univer
sity,
PG
–
ME
–
c
omputer
s
ci
en
ce
and
e
ngineeri
ng
fro
m
Anna
Unive
r
sity.
Compl
et
ed
Ph.D.
in
th
e
Faculty
of
Infor
ma
ti
on
and
Comm
unicati
on
with
spec
ia
l
iz
a
ti
on
in
c
loud
d
ata
s
e
cur
i
ty
Anna
Univer
sity
i
n
Nov
em
be
r
2022.
H
e
has
pu
bli
shed
seve
r
al
r
ese
arc
h
pub
li
c
ati
ons
in
ref
ereed
i
nte
rna
ti
ona
l
j
our
nal
s
wi
th
high
i
mpact
fact
or
and
i
n
te
rna
ti
o
nal
c
onfe
r
ences
as
wel
l.
He
is
th
e
l
ife
m
em
be
r
of
ISTE
and
a
m
em
ber
of
IE
T.
H
e
c
an
b
e con
ta
c
te
d
at e
m
ail: nrrej
inp
aul
@gm
ai
l
.
com.
Varadan
Sh
arm
il
a
is
an
a
ss
ista
nt
p
rofe
ss
or
in
th
e
D
epa
rt
ment
of
CS
E,
R
.
M.
D
Engi
ne
eri
ng
Col
le
ge
,
Chennai
.
Compl
eted
UG
–
BE
c
omputer
s
c
ie
nc
e
and
e
ng
in
ee
ring
fro
m
Anna
Univer
si
t
y,
PG
–
ME
–
c
o
mput
er
s
cienc
e
and
e
ng
ine
e
ri
ng
from
Anna
Univer
sity
.
Purs
uing
Ph.D.
in
th
e
Faculty
o
f
Inform
at
ion
a
nd
Com
munica
ti
on
wi
th
s
pec
i
a
li
z
at
ion
in
m
edica
l
i
m
age
p
roc
essing,
Anna
Univer
sity.
She
has
publi
shed
s
e
ver
al
rese
arc
h
p
ubli
c
at
ions
in
r
efe
re
ed
i
nt
e
rna
ti
on
al
j
ourna
ls
and
i
nt
ern
a
tional
c
onfe
r
ence
s
as
we
ll
.
She
is
the
l
ife
me
mb
er
of
I
STE
and
a
me
mb
er
of
I
ET
.
She
ca
n
be
contac
t
ed
a
t
emai
l:
sharmi
l
ava
rad
h
a
n@gma
il.c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
omp E
ng
IS
S
N:
20
88
-
8708
Imple
men
ti
ng
c
loud c
omp
uting i
n
d
r
ug
d
isc
overy
and
t
el
emed
ic
ine
…
(
P
al
ayanoor
S
eet
hapathy
R
ama
praba
)
1141
Venkatac
halam
Ramesh
Bab
u
cur
ren
tl
y
working
as
a
D
ea
n
of
the
Univer
si
t
y
Journals
and
P
ro
fessor
of
CS
E
a
t
Dr.M.G.
R
.
Edu
ca
t
iona
l
and
R
ese
arc
h
Insti
tut
e
.
He
has
got
27
ye
ars
of
exp
eri
en
ce
in
t
ea
ch
ing
and
4
ye
ars
from
industry
.
He
did
M.E
.
in
c
ompu
ter
s
ci
ence
and
e
ng
ine
er
ing
and
a
furthe
r
Ph.D
in
the
sam
e
disc
ipl
ine.
His
bro
a
d
fi
el
d
of
r
ese
arc
h
was
i
ma
ge
proc
essin
g.
His
ar
ea
of
int
er
ests
include
s
int
ern
et
of
th
ings
(
I
o
T
)
,
m
ac
h
ine
l
e
arn
in
g
and
b
ig
da
ta.
He
has
publi
she
d
rese
arc
h
pap
er
s
in
bo
th
int
e
rna
ti
onal
and
nat
ion
al
journ
als
of
rep
ute.
Be
sides
his
ac
ad
e
mi
c
stin
t,
h
e
h
as
won
ma
ny
a
wards
and
ac
co
la
des
.
He
r
ec
e
ive
d
d
isti
ng
uished
t
e
chnol
o
gy
a
uthor
a
war
d
by
n
ational
t
rai
lblaz
ers
t
rium
ph
a
ward
i
n
the
ye
ar
2023
.
He
has
publ
ished
patent
and
al
so
publi
shed
a
book
on
m
ac
h
ine
l
ea
rn
in
g.
He
a
lso
serve
s
as
the
me
mb
er
of
the
b
o
ard
of
s
tudi
es.
H
e
is
th
e
me
mb
er
of
te
chn
ic
a
l
soci
et
i
es
li
ke
IST
E,
CS
TA,
IAA
C,
IAE
NG
,
ICORS
A.
He
has
orga
ni
zed
works
hops
and
c
onfe
r
ences
a
t
both
na
ti
ona
l
and
int
ern
at
ion
al
le
v
el.
He
has
serve
d
as
a
sess
ion
ch
a
ir
in
conf
ere
n
ce
s.
He
ca
n
b
e cont
a
cted
at email:
r
ameshbabu.
cse
@dr
m
grdu.
ac.i
n
.
Raman
Ramya
is
an
assistant
pro
fessor
i
n
Depa
rt
me
nt
o
f
Elec
tron
ic
s
a
n
d
Comm
unicati
on
Engi
ne
eri
ng
at
Chenna
i
Instit
ut
e
of
Te
chno
logy
,
Kundrat
hur
Ch
enna
i
from
Janua
ry
2024.
She
serve
d
as
a
ss
ista
nt
p
rof
essor
D
epa
rtment
of
E
CE
at Gnanama
ni
C
ollege
o
f
T
ec
hno
logy
fro
m
2010
to
2023
and
oth
er
e
ngin
e
eri
ng
c
olleges.
S
he
has
more
th
a
n
18
y
ea
rs
of
te
a
chi
ng
exp
eri
en
ce.
She
is
doing
her
Ph
.
D.
at
Anna
Univer
sity
,
Chenn
ai
and
her
postgradua
t
e
studie
s
(2008
-
2011
)
a
t
Mahe
ndr
a
E
ngineeri
ng
C
ol
le
ge
,
M
al
l
asa
mu
thra
m
an
d
under
gra
dua
te
(
1996
-
2000)
studie
s
at
Per
iya
r
Maniya
m
m
ai
C
oll
eg
e
of
Engi
n
ee
ring
an
d
Te
chno
logy
for
Wo
me
n,
Tha
n
ja
vur.
Her
r
ese
arc
h
intere
sts
lie
in
the
ar
ea
of
wire
le
ss
com
munica
ti
on,
d
ee
p
l
ea
rn
ing.
R
am
ya
h
as
serve
d
on
roughly
10
+
conf
er
ence
and
works
hop
and
fa
culty
dev
el
opm
ent
progr
am
s.
R
am
y
a
is
the
proud
an
d
obsess
ed
mot
h
er
of
tw
o
c
hil
dr
en,
born
2
001
and
2004
.
S
he
enj
oyed
cook
ing,
li
st
eni
ng
mu
sic
and
rea
d
ing.
She
c
an
be
cont
a
ct
ed
a
t em
a
il
: rram
y
a@c
i
tc
h
enna
i
.
net.
Su
b
biah
Murugan
is
an
a
d
junc
t
p
rofe
ss
or,
Savee
th
a
Schoo
l
of
Engi
n
ee
r
in
g
,
Savee
th
a
Insti
tu
te
of
Medical
a
nd
Techni
c
al
S
ci
en
ce
s,
Chennai,
Ta
m
il
Nadu
,
India
.
He
publi
shed
h
is
re
sea
rch
articles
i
n
ma
ny
interna
tional
and
n
at
ion
a
l
conf
ere
n
ce
s
an
d
journals.
His
rese
arc
h
ar
e
as
inc
l
ude
n
et
w
ork
sec
uri
ty
and
machin
e
learni
ng.
He
ca
n
be
c
onta
c
te
d
at
smuresjur@gm
a
il
.
com
.
Evaluation Warning : The document was created with Spire.PDF for Python.