I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
.
14
, N
o.
4
,
A
ugus
t
2025
, pp.
3003
~
3013
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
14
.i
4
.pp
3003
-
3013
3003
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
In
ve
st
i
gat
i
on
on
l
ow
-
p
e
r
f
or
m
an
c
e
t
u
n
e
d
-
r
e
gr
e
ss
or
of
i
n
h
i
b
i
t
or
y
c
on
c
e
n
t
r
at
i
on
t
a
r
ge
t
i
n
g t
h
e
S
A
R
S
-
C
o
V
-
2
p
ol
yp
r
ot
e
i
n
1ab
D
an
ie
l
F
e
b
r
ia
n
S
e
n
gk
e
y
1,
5
,
7
, A
n
ge
li
n
a S
t
e
van
y R
e
gi
n
a M
as
e
n
gi
2,
5
,
A
lwi
n
M
e
lk
ie
S
am
b
u
l
1,
5
,
T
r
in
a E
k
aw
at
i
T
al
le
i
3,
4,
5
,
S
h
e
r
w
in
R
e
in
al
d
o U
n
s
r
at
d
ia
n
t
o S
om
p
ie
1,
6
1
D
e
pa
r
t
m
e
nt
of
E
l
e
c
t
r
i
c
a
l
E
ngi
ne
e
r
i
ng,
F
a
c
ul
t
y of
E
ngi
ne
e
r
i
ng, U
ni
ve
r
s
i
t
a
s
S
a
m
R
a
t
ul
a
ngi
, M
a
na
do, I
ndone
s
i
a
2
D
e
pa
r
t
m
e
nt
of
P
ha
r
m
a
c
ol
ogy a
nd
T
he
r
a
py
, F
a
c
ul
t
y of
M
e
di
c
i
ne
, U
ni
ve
r
s
i
t
a
s
S
a
m
R
a
t
ul
a
ngi
, M
a
na
do,
I
ndone
s
i
a
3
D
e
pa
r
t
m
e
nt
of
B
i
ol
ogy, F
a
c
ul
t
y of
M
a
t
he
m
a
t
i
c
s
a
nd N
a
t
ur
a
l
S
c
i
e
nc
e
,
U
ni
ve
r
s
i
t
a
s
S
a
m
R
a
t
ul
a
ngi
, M
a
na
do, I
ndone
s
i
a
4
D
e
pa
r
t
m
e
nt
of
B
i
ol
ogy, F
a
c
ul
t
y of
M
e
di
c
i
ne
, U
ni
ve
r
s
i
t
a
s
S
a
m
R
a
t
ul
a
ngi
, M
a
n
a
do, I
ndone
s
i
a
5
B
i
om
ol
e
c
ul
a
r
L
a
bor
a
t
or
y, U
ni
ve
r
s
i
t
a
s
S
a
m
R
a
t
ul
a
ngi
, M
a
na
do, I
ndone
s
i
a
6
I
nf
or
m
a
t
i
on a
nd C
om
m
uni
c
a
t
i
on T
e
c
hnol
ogy A
c
a
de
m
i
c
S
uppor
t
U
ni
t
, U
ni
ve
r
s
i
t
a
s
S
a
m
R
a
t
ul
a
ngi
, M
a
n
a
do, I
ndone
s
i
a
7
D
i
r
e
c
t
or
a
t
e
of
R
e
s
e
a
r
c
h, D
e
ve
l
opm
e
nt
, a
nd I
nnova
t
i
on, I
ndone
s
i
a
A
r
t
i
f
i
c
i
a
l
I
nt
e
l
l
i
ge
nc
e
S
oc
i
e
t
y, J
a
ka
r
t
a
, I
ndone
s
i
a
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
O
c
t
19, 2024
R
e
vi
s
e
d
J
un 12, 2025
A
c
c
e
pt
e
d
J
ul
10, 2025
Hyperparameter
tuning
is
a
key
optimization
strategy
in
machine
le
arning
(ML)
,
often
used
with
GridSearchCV
to
find
optimal
hyperparameter
combinat
ions.
This
study
aimed
to
predict
the
half
-
maximal
inh
ibitory
concentrati
on
(IC
50
)
of
small
molecules
targeting
the
SARS
-
CoV
-
2
re
plicase
polyprotein
1ab
(pp1ab)
by
optimizing
three
ML
algorit
hms:
hist
ogram
gradient
boosting
regressor
(HGBR),
light
gradient
boosting
re
gressor
(LGBR),
and
random
forest
regressor
(RFR).
Bioactivity
data,
inc
luding
duplicates,
were
processed
using
three
approaches:
untreated,
aggrega
tion
of
quantitative
bioactivity,
and
duplicate
removal.
Molecular
featur
es
were
encoded
using
twelve
types
of
molecular
fingerprin
ts.
To
optimi
ze
the
models,
hyperpar
ameter
tuning
with
GridSearc
hCV
was
applied
ac
ross
a
broad
parameter
space.
The
results
showed
that
the
performance
of
the
models
was
inconsistent,
despite
comprehensive
hyperparameter
t
uning.
Further
analys
is
showed
that
the
distribu
tion
of
Murcko
fragments
was
uneven
between
the
training
and
testing
datasets.
Key
fragments
were
underrepresented
in
the
testing
phas
e,
leading
to
a
mismatch
in
model
predictions.
The
study
demonstrates
that
hyperparameter
tuning
alon
e
may
not
be
sufficient
to
achieve
high
predictive
performance
whe
n
the
distribution
of
molecular
fragments
is
unbalanced
between
trainin
g
and
testing
datasets.
Ensuring
fragment
diversity
across
datasets
is
cruc
ial
for
improving mode
l reliability in
drug discov
ery applic
ations.
K
e
y
w
o
r
d
s
:
H
ype
r
pa
r
a
m
e
te
r
t
uni
ng
I
nhi
bi
to
r
y c
onc
e
nt
r
a
ti
on
50
M
ur
c
ko f
r
a
gm
e
nt
s
Q
ua
nt
it
a
ti
ve
s
tr
uc
tu
r
e
-
a
c
ti
vi
ty
r
e
la
ti
ons
hi
p
S
A
R
S
-
C
oV
-
2
p
ol
ypr
ot
e
in
1a
b
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
D
a
ni
e
l
F
e
br
ia
n S
e
ngke
y
D
e
pa
r
tm
e
nt
of
E
le
c
tr
ic
a
l
E
ngi
ne
e
r
in
g,
F
a
c
ul
ty
of
E
ngi
ne
e
r
in
g
,
U
ni
ve
r
s
it
a
s
S
a
m
R
a
tu
la
ngi
K
a
m
pus
U
ns
r
a
t
S
tr
e
e
t
, B
a
hu,
M
a
na
do 95115, I
ndone
s
ia
E
m
a
il
:
da
ni
e
ls
e
ngke
y@
uns
r
a
t.
a
c
.i
d
1.
I
N
T
R
O
D
U
C
T
I
O
N
T
he
C
O
V
I
D
-
19
pa
nde
m
ic
w
a
s
one
of
th
e
f
or
c
e
s
th
a
t
dr
o
v
e
th
e
s
ur
ge
in
c
om
put
e
r
-
a
id
e
d
dr
ug
di
s
c
ove
r
y
(
C
A
D
D
a
dopt
io
n
)
.
R
e
la
te
d
s
tu
di
e
s
dur
in
g
th
is
pe
r
io
d
ta
r
ge
t
e
it
he
r
th
e
hos
t
ta
r
ge
t
,
s
uc
h
a
s
th
e
tr
a
ns
m
e
m
br
a
ne
pr
ot
e
a
s
e
s
e
r
in
e
2
(
T
M
P
R
S
S
2)
[
1]
,
or
th
e
pa
r
t
of
th
e
vi
r
us
,
s
uc
h
a
s
th
e
3
-
c
hym
ot
r
yps
in
-
li
ke
pr
ot
e
a
s
e
(
3C
L
pr
o)
or
th
e
m
a
in
pr
ot
e
a
s
e
(M
pr
o
)
[
2]
–
[
9]
.
H
ua
ng
e
t
al
.
[
1]
ut
il
iz
e
d
m
ol
e
c
ul
a
r
doc
ki
ng
to
e
xa
m
in
e
th
e
dr
ugs
w
it
h
po
s
it
iv
e
ly
c
ha
r
ge
d
gu
a
ni
di
nobe
nz
oy
l
a
nd/
or
a
m
in
id
in
obe
nz
oyl
gr
oups
to
in
hi
bi
t
T
M
P
R
S
S
2
a
t
th
e
hos
t.
M
ol
e
c
ul
a
r
doc
ki
ng
w
a
s
a
ls
o
us
e
d
to
a
s
s
e
s
s
th
e
pot
e
nt
ia
l
to
r
e
pur
pos
e
a
ppr
ove
d
dr
ugs
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
3003
-
3013
3004
s
uc
h
a
s
qui
nol
in
e
[
2]
a
s
w
e
ll
a
s
is
a
vuc
ona
z
oni
um
,
α
-
LI
,
a
nd
pe
nt
a
ga
s
tr
in
[
6]
to
in
hi
bi
t
th
e
vi
r
us
's
m
a
in
pr
ot
e
a
s
e
.
W
hi
le
a
ls
o
t
a
r
ge
ti
ng
th
e
m
a
in
pr
ot
e
a
s
e
,
na
tu
r
a
l
pr
od
uc
ts
w
e
r
e
a
s
s
e
s
s
e
d
w
it
h
m
ol
e
c
ul
a
r
doc
ki
ng
a
s
a
lt
e
r
na
ti
ve
pha
r
m
a
c
ot
he
r
a
py opti
ons
[
4]
, [
8]
, [
9]
.
T
he
a
dopt
io
n
of
m
a
c
hi
ne
le
a
r
ni
ng
(
M
L
)
is
a
va
r
ia
ti
on
in
C
A
D
D
,
known
a
s
m
a
c
hi
ne
le
a
r
ni
ng
-
a
id
e
d
dr
ug
di
s
c
ove
r
y
(
M
L
D
D
)
.
C
la
s
s
if
ic
a
ti
on
is
a
c
om
m
on
ta
s
k
in
M
L
D
D
,
w
he
r
e
t
he
ta
r
ge
t
of
th
e
c
la
s
s
if
ic
a
ti
on
us
e
s
e
it
he
r
a
known
in
te
r
a
c
ti
on,
c
ode
d
a
s
a
bi
n
a
r
y
va
lu
e
[
10
]
–
[
12]
or
c
a
te
gor
ie
s
ba
s
e
d
on
th
e
di
s
c
r
e
ti
z
e
d
ha
lf
-
m
a
xi
m
a
l
in
hi
bi
to
r
y
c
onc
e
nt
r
a
ti
on
(
I
C
50
)
va
lu
e
[
7]
,
[
13]
,
[
1
4]
.
D
e
s
pi
te
th
e
di
s
c
r
e
ti
z
a
ti
on
of
th
e
I
C
50
be
in
g
a
c
om
m
on
a
ppr
oa
c
h
a
s
de
m
on
s
tr
a
te
d
in
th
e
m
e
nt
io
ne
d
s
tu
di
e
s
,
how
e
ve
r
,
th
is
a
ppr
oa
c
h
i
s
di
s
c
our
a
ge
d
in
ge
ne
r
a
l
e
pi
de
m
io
lo
gy
s
tu
di
e
s
du
e
to
th
e
lo
s
s
of
in
f
or
m
a
ti
on
w
it
hi
n
th
e
num
e
r
ic
va
r
ia
bl
e
[
15]
.
B
a
s
e
d
on
two
m
e
ta
-
a
na
ly
s
e
s
,
it
is
f
ound
th
a
t
c
ont
in
uou
s
,
r
a
th
e
r
th
a
n
di
s
c
r
e
t
e
,
m
e
a
s
ur
e
m
e
nt
s
c
oul
d
im
pr
ove
va
li
di
ty
a
nd
r
e
li
a
bi
li
ty
[
16]
.
F
or
in
s
ta
nc
e
,
G
a
o
e
t
al
.
[
17]
bui
ld
r
e
gr
e
s
s
io
n
m
ode
ls
us
in
g
r
a
ndom
f
or
e
s
t
(
RF
)
,
a
nd
s
uppor
t
ve
c
to
r
m
a
c
hi
ne
(
S
V
M
)
w
it
h
s
om
e
opt
im
iz
a
ti
on
to
pr
e
di
c
t
th
e
I
C
50
of
th
e
[
1,2,3]
tr
ia
z
ol
o
[
4,5
-
d]
pyr
i
m
id
in
e
de
r
iv
a
ti
ve
s
(
1,2,3
-
T
P
D
)
to
in
hi
bi
t
th
e
r
e
pl
ic
a
ti
on
of
th
e
M
G
C
-
803,
th
e
ga
s
tr
ic
c
a
nc
e
r
c
e
ll
in
hum
a
n
s
.
I
n
c
ont
r
a
s
t
to
th
e
w
or
k
in
[
13]
,
[
18]
ut
il
iz
e
d
S
V
M
,
a
r
ti
f
ic
ia
l
ne
ur
a
l
ne
twor
k
(
A
N
N
)
,
k
ne
a
r
e
s
t
ne
ig
hbor
(
K
NN)
,
a
nd
R
F
to
bui
ld
r
e
gr
e
s
s
io
n
m
ode
ls
f
or
p
r
e
di
c
ti
ng
th
e
I
C
50
to
w
a
r
ds
m
ul
ti
pl
e
h
e
pa
ti
ti
s
C
vi
r
us
(
H
C
V
)
non
-
s
tr
u
c
tu
r
a
l
pr
ot
e
in
s
.
S
im
il
a
r
ly
,
th
e
w
or
k
of
F
i
a
t
e
t
al
.
[
1
9]
us
e
d
r
a
ndom
f
or
e
s
t
r
e
gr
e
s
s
i
on
(
R
F
R
)
a
nd
gr
a
di
e
n
t
boo
s
ti
n
g
r
e
gr
e
s
s
io
n
(
G
B
R
)
to
d
e
ve
lo
p M
L
m
od
e
ls
to
p
r
e
di
c
t
t
he
I
C
50
,
t
a
r
ge
ti
ng
th
e
H
C
V
ge
no
ty
pe
1a
(
I
s
ol
a
te
1)
.
S
uppor
t
ve
c
to
r
r
e
gr
e
s
s
io
n
(
S
V
R
)
is
us
e
d
in
pr
e
di
c
ti
ng
th
e
in
hi
bi
ti
on
of
s
m
a
ll
m
ol
e
c
ul
e
s
to
be
ta
-
s
e
c
r
e
ta
s
e
1
(
B
A
C
E
1)
,
w
hi
c
h
is
a
n
e
nz
ym
e
r
e
la
te
d
to
A
lz
he
im
e
r
’
s
di
s
e
a
s
e
(
A
D
)
[
20]
.
I
n
a
not
he
r
s
tu
dy
,
m
ul
ti
pl
e
li
ne
a
r
r
e
gr
e
s
s
io
ns
(
M
L
R
)
w
as
f
ound
a
s
th
e
b
e
s
t
a
lg
or
it
hm
c
om
pa
r
e
d
to
S
V
R
,
c
la
s
s
if
ic
a
ti
on
a
nd
r
e
gr
e
s
s
io
n
(
C
A
R
T
)
,
a
nd
A
N
N
in
pr
e
di
c
ti
ng
th
e
c
om
pound
bi
ndi
ng
f
r
e
e
e
ne
r
gy
(
B
F
E
)
to
w
a
r
ds
th
e
S
A
R
S
-
C
oV
-
2 m
a
in
pr
ot
e
a
s
e
[
21]
.
I
n our
p
r
e
vi
ous
s
tu
dy
[
22]
,
we
e
xpe
r
im
e
nt
e
d w
it
h 42
M
L
r
e
gr
e
s
s
io
n a
lg
or
it
hm
s
t
o pr
e
di
c
t
th
e
I
C
50
o
f
bi
oa
c
ti
ve
c
om
pounds
,
t
a
r
ge
ti
ng t
he
pol
ypr
ot
e
in
1a
b
(
pp1a
b
)
of
t
he
S
A
R
S
-
C
oV
-
2, w
hi
c
h c
om
pr
is
e
s
t
he
vi
r
us
’
s
non
-
s
tr
uc
tu
r
a
l
pr
ot
e
in
(
N
S
P
)
12
to
N
S
P
16
[
23]
,
[
24]
.
T
he
de
f
a
ul
t
hype
r
pa
r
a
m
e
te
r
s
w
e
r
e
us
e
d
w
it
hout
a
ny
tu
ni
ng
pr
oc
e
s
s
in
vol
ve
d.
T
he
f
e
a
tu
r
e
s
w
e
r
e
de
r
iv
e
d
f
r
om
th
e
c
om
pounds
by
us
in
g
P
ubC
h
e
m
f
in
ge
r
pr
in
ts
.
O
ut
of
th
e
42
e
xpe
r
im
e
nt
e
d
a
lg
or
it
hm
s
,
th
r
e
e
a
lg
or
it
hm
s
:
R
F
R
,
li
ght
gr
a
di
e
nt
boos
ti
ng
m
a
c
hi
ne
r
e
gr
e
s
s
io
n
(
L
G
B
R
)
,
a
nd
hi
s
to
gr
a
m
gr
a
di
e
nt
boos
ti
ng
m
a
c
hi
ne
r
e
gr
e
s
s
io
n
(
H
G
B
R
)
w
e
r
e
f
ound
a
s
th
e
m
os
t
s
t
a
bl
e
f
or
th
is
c
om
bi
na
ti
on
ba
s
e
d
on
th
e
R
2
va
lu
e
s
.
H
ype
r
pa
r
a
m
e
te
r
tu
ni
ng
i
s
a
te
c
hni
que
in
M
L
th
a
t
is
us
e
d
to
opt
im
iz
e
th
e
m
ode
l
pe
r
f
or
m
a
nc
e
by
twe
a
ki
ng
th
e
hype
r
pa
r
a
m
e
te
r
s
of
th
e
a
l
gor
it
hm
[
25
]
–
[
28]
.
I
t
is
c
om
m
onl
y
us
e
d
w
it
h
G
r
id
S
e
a
r
c
hC
V
,
w
hi
c
h
c
om
bi
ne
s
a
la
r
ge
hyp
e
r
pa
r
a
m
e
te
r
s
e
a
r
c
h
s
pa
c
e
a
nd
c
r
os
s
-
va
li
da
ti
on
to
obt
a
in
th
e
opt
im
a
l
ge
ne
r
a
li
z
a
bl
e
m
ode
l
f
or
th
e
a
lg
or
it
hm
.
T
he
r
e
f
or
e
,
in
th
is
s
tu
dy,
w
e
e
xt
e
nde
d
th
e
e
xp
e
r
im
e
nt
w
it
h
th
e
s
e
a
lg
or
it
hm
s
,
w
hi
c
h
a
ls
o
f
a
ll
in
to
th
e
e
n
s
e
m
bl
e
tr
e
e
-
ba
s
e
d
c
a
te
gor
y,
a
nd
in
ve
s
ti
ga
te
d
th
e
im
pa
c
ts
of
da
ta
di
s
tr
ib
ut
io
n, e
s
pe
c
ia
ll
y t
he
M
ur
c
ko f
r
a
gm
e
nt
s
of
t
he
c
om
pound
s
,
on
th
e
m
ode
l
pe
r
f
or
m
a
nc
e
.
T
he
r
e
s
t
of
th
is
a
r
ti
c
le
is
or
ga
ni
z
e
d
a
s
f
ol
lo
w
s
:
in
s
e
c
ti
on
2
,
w
e
pr
e
s
e
nt
th
e
da
ta
s
e
t
a
s
w
e
ll
a
s
th
e
m
e
th
ods
w
e
u
s
e
d
f
or
da
ta
c
ur
a
ti
on,
tr
e
a
tm
e
nt
s
in
pr
e
-
pr
oc
e
s
s
in
g,
m
ode
l
tr
a
in
in
g,
va
li
da
ti
on,
a
nd
pe
r
f
or
m
a
nc
e
e
va
lu
a
ti
on. T
he
n, i
n
s
e
c
ti
on
3
, w
e
c
om
pa
r
e
t
he
pe
r
f
or
m
a
nc
e
be
twe
e
n t
he
t
r
e
a
tm
e
nt
s
, a
s
w
e
ll
a
s
i
nve
s
ti
ga
te
t
he
di
s
tr
ib
ut
io
n
of
c
om
pound
c
ha
r
a
c
te
r
is
ti
c
s
in
tr
a
in
in
g
a
nd
te
s
ti
ng
da
ta
s
e
ts
.
L
a
s
t,
in
s
e
c
ti
on
4
,
th
is
pa
pe
r
is
c
onc
lu
de
d
,
a
nd dir
e
c
ti
ons
f
or
f
ut
ur
e
w
or
k
a
r
e
pr
e
s
e
nt
e
d.
2.
M
E
T
H
O
D
T
he
r
e
s
e
a
r
c
h
m
e
th
odol
ogy
m
a
in
ly
f
ol
lo
w
s
th
e
c
or
e
a
c
ti
vi
ti
e
s
of
da
ta
s
c
ie
nc
e
m
e
th
odol
ogy
,
a
s
s
how
n
in
F
ig
ur
e
1
m
a
in
ly
c
ons
is
ts
of
th
r
e
e
pa
r
ts
.
T
he
pr
e
pr
oc
e
s
s
in
g
pa
r
t
is
r
e
la
te
d
to
f
a
s
hi
oni
ng
th
e
c
om
pounds
'
bi
oa
c
ti
vi
ty
da
ta
f
or
M
L
tr
a
in
in
g.
T
he
pi
pe
li
ne
pa
r
t
i
s
w
he
r
e
w
e
us
e
c
u
s
to
m
pi
pe
li
ne
s
th
a
t
f
e
e
d
in
to
th
e
hype
r
pa
r
a
m
e
te
r
tu
ni
ng
pr
oc
e
s
s
.
T
h
e
pi
pe
li
ne
d
a
ppr
oa
c
h
w
il
l
e
n
s
ur
e
no
da
ta
le
a
ka
g
e
,
he
nc
e
gua
r
a
nt
e
e
in
g
th
a
t
th
e
m
ode
l
ha
s
ne
ve
r
s
e
e
n
th
e
da
ta
u
s
e
d
f
or
it
s
pe
r
f
or
m
a
nc
e
e
va
lu
a
ti
on.
L
a
s
t,
in
th
e
r
e
s
ul
t
a
na
ly
s
is
a
nd
doc
um
e
nt
a
ti
on
pa
r
t
,
th
e
e
xpe
r
im
e
nt
r
e
s
ul
t
s
a
r
e
a
n
a
ly
zed
a
nd
c
o
m
pa
r
e
d.
M
a
in
ly
,
w
e
u
s
e
d
P
yt
hon
ve
r
s
io
n
3.10
a
nd
s
c
ik
it
-
le
a
r
n
[
29]
ve
r
s
io
n 1.5.1 in t
he
m
ode
li
ng a
nd a
na
ly
s
is
pha
s
e
s
.
2.1
.
P
r
e
p
r
oc
e
s
s
in
g
T
he
da
t
a
pr
e
pa
r
a
ti
on
pha
s
e
b
e
gi
ns
w
it
h
da
ta
a
c
qui
s
it
io
n
,
s
p
e
c
if
ic
a
ll
y
in
hi
bi
to
r
y
bi
oa
c
ti
vi
ty
da
ta
.
B
y
us
in
g
th
e
C
hE
M
B
L
w
e
b
s
e
r
vi
c
e
[
30]
,
w
e
a
c
qui
r
e
d
,
in
to
ta
l,
1,455
c
om
pounds
w
it
h
known
I
C
50
to
th
e
S
A
R
S
-
C
oV
-
2
pp1a
b
(
C
H
E
M
B
L
4523582)
,
he
a
vi
ly
in
c
r
e
a
s
e
d
f
r
om
our
pr
e
vi
ous
s
tu
dy
[
22]
.
I
n
th
is
da
ta
s
e
t,
c
om
pounds
a
r
e
r
e
pr
e
s
e
nt
e
d
in
s
im
pl
if
ie
d
m
ol
e
c
ul
a
r
in
put
li
n
e
e
nt
r
y
s
ys
te
m
(
S
M
I
L
E
S
)
f
or
m
a
t
.
T
he
da
ta
c
le
a
ni
ng
a
l
s
o
in
c
lu
de
s
s
ta
nda
r
di
z
in
g
th
e
S
M
I
L
E
S
not
a
ti
on
of
e
a
c
h
c
om
pound
a
nd
c
onve
r
ti
ng
th
e
I
C
50
to
th
e
r
e
s
pe
c
ti
ve
ne
ga
ti
ve
lo
g
a
r
it
hm
ic
s
c
a
le
,
pI
C
50
,
he
n
c
e
na
r
r
ow
in
g
th
e
s
c
a
le
.
F
ol
lo
w
in
g
th
e
c
le
a
ni
ng
s
te
ps
,
w
e
c
ont
in
ue
w
it
h
tr
e
a
ti
ng
th
e
dupl
ic
a
t
e
s
.
I
n
dr
ug
di
s
c
ove
r
y
e
xp
e
r
im
e
nt
s
,
di
f
f
e
r
e
nt
a
ppr
oa
c
he
s
a
nd
di
f
f
e
r
e
nt
la
bor
a
to
r
y
s
e
tt
in
gs
m
ig
ht
yi
e
ld
di
f
f
e
r
e
nt
I
C
50
va
lu
e
s
,
de
s
pi
te
th
e
us
e
of
th
e
s
a
m
e
c
om
pound.
I
n
our
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
I
nv
e
s
ti
gat
io
n on low
-
pe
r
fo
r
m
anc
e
t
une
d
-
r
e
g
r
e
s
s
or
of
i
nhi
bi
to
r
y
c
onc
e
nt
r
at
io
n …
(
D
ani
e
l
F
e
br
ia
n Se
ng
k
e
y
)
3005
e
xpe
r
im
e
nt
s
,
w
e
tr
ie
d
s
e
ve
r
a
l
a
ppr
oa
c
he
s
to
ha
ndl
e
th
e
dupl
ic
a
te
d
da
ta
.
F
ir
s
t,
w
e
le
f
t
th
e
m
a
s
is
;
s
e
c
ond,
w
e
a
ggr
e
ga
te
d t
he
m
by t
a
ki
ng t
he
a
ve
r
a
ge
of
t
he
pI
C
50
va
lu
e
;
a
nd l
a
s
t,
w
e
dr
oppe
d a
ll
dupli
c
a
te
d
c
om
pounds
.
F
ig
ur
e
1.
C
our
s
e
of
r
e
s
e
a
r
c
h
A
f
te
r
th
e
dupl
ic
a
te
s
w
e
r
e
tr
e
a
t
e
d,
w
e
c
ont
in
ue
d
by
tr
a
ns
f
or
m
in
g
th
e
c
he
m
ic
a
l
c
om
pounds
(
in
S
M
I
L
E
S
)
in
to
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
ts
(
de
s
c
r
ip
to
r
s
)
,
r
e
s
ul
ti
n
g
in
a
ta
bl
e
f
or
e
a
c
h
f
in
ge
r
pr
in
t
w
e
us
e
d.
T
he
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
ts
r
e
pr
e
s
e
nt
th
e
c
ha
r
a
c
te
r
is
ti
c
s
of
a
c
he
m
ic
a
l
c
om
pound.
F
or
e
a
c
h
c
om
pound,
a
f
in
ge
r
pr
in
t
is
a
s
e
r
ie
s
of
bi
ts
, w
he
r
e
e
a
c
h bi
t
is
a
B
ool
e
a
n
, r
e
pr
e
s
e
nt
in
g a
s
pe
c
if
ic
c
he
m
ic
a
l
c
h
a
r
a
c
te
r
is
ti
c
, a
nd,
a
s
a
w
hol
e
,
de
s
c
r
ib
e
s
th
e
c
om
pound.
F
or
in
s
ta
nc
e
,
th
e
P
ubC
he
m
f
in
ge
r
pr
in
t,
th
e
f
ir
s
t
bi
t
s
how
s
w
he
th
e
r
th
e
r
e
s
pe
c
ti
ve
c
om
pound
po
s
s
e
s
s
e
s
f
our
or
m
or
e
hydr
oge
n
a
to
m
s
.
T
he
tr
a
ns
f
or
m
a
ti
ons
to
th
e
f
in
ge
r
pr
in
ts
a
r
e
done
us
in
g
P
a
D
E
L
s
of
twa
r
e
[
31]
.
I
n
to
ta
l,
th
e
r
e
a
r
e
12
va
r
i
a
nt
s
of
f
e
a
tu
r
e
s
e
ts
.
T
h
e
de
s
c
r
ip
ti
on
of
e
a
c
h
f
in
ge
r
pr
in
t
a
nd
th
e
num
be
r
of
m
ol
e
c
ul
a
r
f
e
a
tu
r
e
s
it
ha
s
a
r
e
pr
ovi
de
d
in
T
a
bl
e
1.
S
in
c
e
12
ty
pe
s
of
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
ts
a
r
e
in
us
e
a
nd
th
r
e
e
tr
e
a
tm
e
nt
s
f
or
dupl
ic
a
te
s
,
36
da
ta
s
e
ts
a
r
e
us
e
d
f
or
th
e
e
xpe
r
im
e
nt
s
.
T
he
n,
us
in
g a
n 80:20 r
a
ti
o of
t
r
a
in
in
g a
nd t
e
s
ti
ng da
ta
, r
e
s
pe
c
ti
ve
ly
, e
a
c
h da
ta
s
e
t
is
s
pl
it
us
in
g t
he
f
unc
ti
on a
va
il
a
bl
e
in
s
c
ik
it
-
le
a
r
n
.
2.2
.
P
ip
e
li
n
e
an
d
h
yp
e
r
p
ar
am
e
t
e
r
t
u
n
in
g
T
o
e
ns
ur
e
th
e
r
e
li
a
bi
li
ty
a
nd
th
e
c
ont
in
ui
ty
of
m
ode
l
tr
a
in
in
g
a
nd,
la
te
r
,
ut
il
iz
e
th
e
m
f
or
in
f
e
r
e
nc
in
g,
th
e
f
e
a
tu
r
e
s
e
le
c
ti
on pr
oc
e
s
s
e
s
a
r
e
c
oupl
e
d w
it
h t
he
r
e
gr
e
s
s
or
s
a
s
pi
pe
li
ne
s
. T
he
f
ir
s
t
f
e
a
tu
r
e
s
e
le
c
ti
on me
th
od
is
th
e
va
r
ia
nc
e
th
r
e
s
hol
d
.
T
hi
s
f
e
a
tu
r
e
s
e
le
c
ti
on
m
e
th
od
dr
op
s
f
e
a
tu
r
e
s
w
it
h
va
r
ia
nc
e
unde
r
th
e
s
pe
c
if
ie
d
le
ve
l.
T
he
r
e
s
t
of
th
e
f
e
a
tu
r
e
s
a
r
e
th
e
n
f
e
d
in
to
th
e
s
e
c
ond
f
e
a
t
ur
e
s
e
le
c
ti
on
m
e
th
od,
th
e
m
ut
ua
l
in
f
or
m
a
ti
on
(
e
nt
r
opy
)
. W
e
s
e
t
th
e
f
e
a
tu
r
e
s
s
e
le
c
to
r
t
o us
e
onl
y t
he
t
op c
e
r
ta
in
pe
r
c
e
nt
il
e
, a
c
c
or
di
ng t
o t
he
f
e
a
tu
r
e
s
'
e
nt
r
opy
s
c
or
e
.
T
he
pos
t
-
f
e
a
tu
r
e
s
e
le
c
ti
on
da
ta
s
e
t
w
il
l
th
e
n
be
us
e
d
to
t
r
a
in
th
e
r
e
gr
e
s
s
or
.
A
s
de
s
c
r
ib
e
d
e
a
r
li
e
r
,
th
r
e
e
M
L
r
e
gr
e
s
s
io
n a
lg
or
it
hm
s
w
e
r
e
e
xpl
or
e
d a
lt
e
r
na
te
ly
:
H
G
B
R
, R
F
R
, a
nd L
G
B
R
.
A
s
a
de
v
e
lo
pm
e
nt
f
r
om
our
pr
e
vi
ous
a
ppr
oa
c
h
in
[
22]
,
th
e
c
ur
r
e
nt
m
e
th
od
e
m
pl
oys
hyp
e
r
pa
r
a
m
e
te
r
tu
ni
ng
us
in
g
G
r
id
S
e
a
r
c
hC
V
,
to
e
xh
a
us
ti
ve
ly
t
e
s
t
e
a
c
h
c
om
bi
na
ti
on
of
th
e
hype
r
pa
r
a
m
e
t
e
r
s
in
th
e
s
e
a
r
c
h
s
pa
c
e
.
T
o e
ns
ur
e
t
he
g
e
ne
r
a
li
z
a
bi
li
ty
of
t
he
hype
r
pa
r
a
m
e
te
r
s
w
i
th
t
he
be
s
t
pe
r
f
or
m
a
nc
e
dur
in
g t
r
a
in
in
g, 5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on
is
us
e
d.
T
a
bl
e
1
li
s
ts
th
e
s
te
ps
a
nd
m
odul
e
s
in
th
e
pi
pe
li
ne
s
,
a
nd
th
e
s
e
a
r
c
h
s
pa
c
e
us
e
d
f
or
hype
r
pa
r
a
m
e
te
r
t
uni
ng.
2.3
.
A
n
al
ys
is
an
d
d
oc
u
m
e
n
t
at
io
n
I
n
th
is
pa
r
t
of
th
e
r
e
s
e
a
r
c
h,
w
e
e
va
lu
a
te
th
e
pe
r
f
or
m
a
nc
e
of
th
e
m
ode
ls
by
c
om
pa
r
in
g
th
e
pe
r
f
or
m
a
nc
e
of
th
e
tr
a
in
e
d
m
ode
l
s
a
nd
a
ppl
yi
ng
it
to
in
f
e
r
t
he
la
be
ls
in
th
e
t
e
s
ti
ng
da
t
a
s
e
t.
P
e
r
f
or
m
a
nc
e
m
e
tr
ic
s
us
e
d
a
r
e
R
2
a
nd
th
e
r
oot
m
e
a
n
s
qu
a
r
e
d
e
r
r
or
(
R
M
S
E
)
.
S
ta
ti
s
ti
c
a
l
a
na
ly
s
e
s
a
nd
f
ig
ur
e
s
a
r
e
done
us
in
g
th
e
R
s
ta
ti
s
ti
c
a
l
s
of
twa
r
e
ve
r
s
io
n 4.4.1
[
32]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
3003
-
3013
3006
T
a
bl
e
1
. H
ype
r
pa
r
a
m
e
te
r
t
uni
ng pipe
li
ne
s
t
e
ps
, m
odul
e
, a
nd hy
pe
r
pa
r
a
m
e
te
r
s
e
a
r
c
h s
p
a
c
e
P
i
pe
l
i
ne
s
t
e
p
M
odul
e
H
ype
r
pa
r
a
m
e
t
e
r
s
e
a
r
c
h s
pa
c
e
F
e
a
t
ur
e
s
e
l
e
c
t
i
on
V
a
r
i
a
nc
e
t
hr
e
s
hol
d
t
hr
e
s
hol
d:
0.8(
1
-
0.8)
=0.16; 0.9(
1
-
0.9)
=0.09
S
e
l
e
c
t
pe
r
c
e
nt
i
l
e
(
s
c
or
i
ng by
m
ut
ua
l
i
nf
or
m
a
t
i
on
)
p
e
r
c
e
nt
i
l
e
:
10, 20, 50, 100
R
e
gr
e
s
s
or
H
G
B
R
m
a
x_i
t
e
r
:
[
100, 1000, 10000, 99999999]
,
m
a
x_de
pt
h:
[
N
one
, 10, 20, 30,
40, 50]
,
m
i
n_s
a
m
pl
e
s
_l
e
a
f
:
[
1, 2, 4, 8, 16, 32, 64]
,
l
2_r
e
gul
a
r
i
z
a
t
i
on:
[
0, 0.1, 0.01]
,
l
e
a
r
ni
ng_r
a
t
e
:
[
0.01]
,
w
a
r
m
_s
t
a
r
t
:
[
T
r
ue
, F
a
l
s
e
]
,
e
a
r
l
y_s
t
oppi
ng:
[
T
r
ue
]
,
n_i
t
e
r
_no_c
ha
nge
:
[
10, 100]
,
r
a
ndom
_s
t
a
t
e
:
[
22]
,
L
G
B
R
boos
t
i
ng_t
ype
'
:
[
'
gbdt
'
,'
r
f
'
]
,
n_e
s
t
i
m
a
t
or
s
'
:
[
99999999]
,
m
a
x_de
pt
h'
:
[
-
1, 15, 31, 63]
,
l
e
a
r
ni
ng_r
a
t
e
:
[
0.01]
,
r
a
ndom
_s
t
a
t
e
:
[
22]
,
num
_l
e
a
ve
s
:
[
7, 31, 127, 1027, 2047, 4095]
e
a
r
l
y_s
t
oppi
ng_r
ounds
:
[
20]
R
F
R
n_e
s
t
i
m
a
t
or
s
:
[
10, 100, 1000]
,
m
i
n_s
a
m
pl
e
s
_l
e
a
f
:
[
1, 2, 4, 8, 16]
,
m
a
x_de
pt
h:
[
N
one
, 10, 20, 30, 40, 50]
,
oob_s
c
or
e
:
[
T
r
ue
, F
a
l
s
e
]
,
r
a
ndom
_s
t
a
t
e
:
[
22]
,
w
a
r
m
_s
t
a
r
t
:
[
T
r
ue
, F
a
l
s
e
]
,
m
i
n_s
a
m
pl
e
s
_s
pl
i
t
:
[
2, 3, 4, 8, 16]
,
m
a
x_f
e
a
t
ur
e
s
:
[
"
s
qr
t
", "
l
og2", N
one
]
3.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
U
s
in
g
th
e
be
s
t
hype
r
p
a
r
a
m
e
te
r
s
f
o
und
f
or
e
a
c
h
c
om
bi
na
ti
o
n
of
m
ol
e
c
u
la
r
f
i
nge
r
pr
i
nt
s
a
n
d
a
lg
or
it
h
m
f
or
e
v
e
r
y
tr
e
a
tm
e
nt
of
du
pl
ic
a
te
d
d
a
ta
,
w
e
tr
a
in
e
d
m
o
de
l
s
a
nd
a
ppl
ie
d
th
e
m
to
th
e
t
e
s
t
in
g da
t
a
s
e
t.
T
a
bl
e
s
2
t
o
4
s
how
th
e
be
s
t
hype
r
pa
r
a
m
e
te
r
s
f
or
H
G
B
R
,
L
G
B
R
,
a
nd
R
F
R
a
lg
or
it
hm
s
,
r
e
s
pe
c
ti
ve
ly
,
f
or
e
a
c
h
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
t.
3.1
.
P
e
r
f
or
m
an
c
e
m
e
t
r
ic
s
F
ig
ur
e
2
s
how
s
t
he
boxplot
s
of
th
e
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
,
R
2
,
a
nd
R
M
S
E
of
th
e
m
ode
ls
w
it
h
th
e
hype
r
pa
r
a
m
e
te
r
s
th
a
t
ga
ve
th
e
be
s
t
pe
r
f
or
m
a
nc
e
dur
in
g
th
e
tu
n
in
g
w
it
h
th
e
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on
s
te
p.
I
t
is
obvi
ous
th
a
t
by
e
nt
ir
e
ly
dr
oppi
ng
th
e
dupl
ic
a
te
d
bi
oa
c
ti
vi
ty
da
t
a
,
th
e
m
ode
ls
pe
r
f
or
m
e
d
e
xt
r
e
m
e
ly
di
f
f
e
r
e
nt
ly
f
r
om
th
e
ot
he
r
two
tr
e
a
tm
e
nt
s
.
W
he
n
th
e
dupl
ic
a
te
d
da
ta
w
a
s
dr
oppe
d,
th
e
R
2
va
lu
e
s
dr
oppe
d
a
nd
c
om
m
onl
y
f
e
ll
unde
r
z
e
r
o
w
it
h
a
hi
ghe
r
va
r
ia
ti
on,
e
it
he
r
dur
in
g
tr
a
in
in
g
or
te
s
ti
ng,
a
s
c
a
n
b
e
s
e
e
n
in
F
ig
ur
e
2
(
a
)
.
M
e
a
nw
hi
le
,
w
h
e
n
th
e
s
e
dupl
ic
a
t
e
s
w
e
r
e
le
f
t
unt
ouc
he
d,
th
e
R
2
dur
in
g
tr
a
in
in
g
w
a
s
s
li
ght
ly
lo
w
e
r
th
a
n
th
e
a
ve
r
a
ge
d
pI
C
50
tr
e
a
tm
e
nt
,
but
th
e
c
ondi
ti
on
w
a
s
r
e
ve
r
s
e
d
in
te
s
ti
ng.
T
he
boxplot
of
th
e
lo
s
s
f
unc
ti
on,
R
M
S
E
,
in
F
ig
ur
e
2
(
b
)
in
di
c
a
te
s
th
e
s
a
m
e
th
in
g.
P
e
r
f
or
m
a
nc
e
m
e
tr
ic
s
f
or
e
a
c
h
c
om
bi
na
ti
on
of
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
t
a
nd a
lg
or
it
hm
w
it
h t
he
be
s
t
hype
r
pa
r
a
m
e
te
r
s
a
r
e
s
how
n i
n F
ig
u
r
e
s
1 a
nd 2.
(
a
)
(
b)
F
ig
ur
e
2. B
oxpl
ot
s
of
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
a
c
r
os
s
t
r
e
a
tm
e
nt
s
a
n
d m
ode
li
ng s
ta
ge
s
of
(
a
)
R
2
a
nd (
b)
R
M
S
E
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
I
nv
e
s
ti
gat
io
n on low
-
pe
r
fo
r
m
anc
e
t
une
d
-
r
e
g
r
e
s
s
or
of
i
nhi
bi
to
r
y
c
onc
e
nt
r
at
io
n …
(
D
ani
e
l
F
e
br
ia
n Se
ng
k
e
y
)
3007
B
e
f
or
e
s
ta
ti
s
ti
c
a
ll
y
c
om
pa
r
in
g
th
e
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
,
th
e
S
ha
pi
r
o
-
W
il
k
te
s
t
w
a
s
a
ppl
ie
d
to
c
he
c
k
th
e
di
s
tr
ib
ut
io
n
nor
m
a
li
ty
of
e
a
c
h
pe
r
f
or
m
a
nc
e
m
e
tr
ic
.
F
or
th
is
te
s
t,
th
e
da
ta
a
r
e
gr
oupe
d
a
c
c
or
di
ng
to
tr
e
a
tm
e
nt
s
,
a
lg
or
it
hm
s
,
a
nd
m
ode
li
ng
s
ta
ge
s
.
T
he
r
e
f
or
e
,
a
s
in
gl
e
di
s
tr
ib
ut
io
n
te
s
te
d
h
a
s
12
p
e
r
f
or
m
a
nc
e
da
ta
.
T
a
bl
e
2
s
how
s
th
e
p
-
va
lu
e
s
of
th
e
S
ha
pi
r
o
-
W
il
k
te
s
t.
W
it
h
α
=
0.05,
it
is
c
le
a
r
th
a
t
s
om
e
of
th
e
d
a
ta
a
r
e
not
nor
m
a
ll
y di
s
tr
ib
ut
e
d, he
nc
e
non
-
pa
r
a
m
e
tr
ic
t
e
s
t
s
houl
d be
us
e
d
f
or
f
ur
th
e
r
a
na
ly
s
is
.
T
a
bl
e
2
. P
-
va
lu
e
s
of
t
he
S
ha
pi
r
o
-
W
il
k t
e
s
t
f
or
nor
m
a
li
ty
di
s
tr
ib
ut
io
n of
t
he
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
, gr
oupe
d by
th
e
t
r
e
a
tm
e
nt
f
or
dupli
c
a
te
s
a
nd a
lg
or
it
hm
s
. T
he
i
ta
li
c
iz
e
d num
be
r
s
a
r
e
t
hos
e
unde
r
th
e
α
=
0.05
T
r
e
a
t
m
e
nt
A
l
gor
i
t
hm
T
r
a
i
n R
2
T
e
s
t
R
2
T
r
a
i
n R
M
S
E
T
e
s
t
R
M
S
E
U
nt
r
e
a
t
e
d
H
G
B
R
0.038
0.167
0.083
0.124
U
nt
r
e
a
t
e
d
R
F
R
0.017
0.197
0.047
0.035
U
nt
r
e
a
t
e
d
L
G
B
R
0.018
0.202
0.051
0.036
A
ve
r
a
ge
d
H
G
B
R
0.017
0.574
0.089
0.879
A
ve
r
a
ge
d
R
F
R
0.012
0.475
0.047
0.842
A
ve
r
a
ge
d
L
G
B
R
0.012
0.360
0.048
0.681
D
r
oppe
d
H
G
B
R
0.001
<0.001
0.841
0.205
D
r
oppe
d
R
F
R
0.255
0.537
0.210
0.649
D
r
oppe
d
L
G
B
R
0.282
0.629
0.132
0.812
W
e
us
e
d
th
e
F
r
ie
dm
a
n
te
s
t
f
or
one
-
w
a
y
r
e
pe
a
te
d
m
e
a
s
ur
e
s
a
na
ly
s
is
of
v
a
r
ia
nc
e
to
c
om
pa
r
e
e
a
c
h
pe
r
f
or
m
a
nc
e
m
e
tr
ic
be
twe
e
n
th
e
tr
e
a
tm
e
nt
s
w
it
h
th
e
s
a
m
e
a
lg
or
it
hm
,
by
us
in
g
th
e
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
t
a
s
th
e
id
e
nt
if
ie
r
.
T
he
r
e
s
ul
ts
,
a
s
s
how
n
in
T
a
bl
e
3
,
s
how
th
a
t
in
a
l
l
c
om
pa
r
is
ons
,
a
t
le
a
s
t
one
gr
oup
of
dupl
ic
a
te
da
ta
t
r
e
a
tm
e
nt
ha
s
a
s
ig
ni
f
ic
a
nt
ly
di
f
f
e
r
e
nt
di
s
tr
ib
ut
io
n
of
a
pa
r
t
ic
ul
a
r
pe
r
f
or
m
a
nc
e
m
e
tr
ic
.
F
ol
lo
w
in
g t
he
one
-
w
a
y
r
e
pe
a
te
d
m
e
a
s
ur
e
s
F
r
ie
dm
a
n
te
s
t,
w
e
c
a
r
r
ie
d
out
th
e
P
a
i
r
w
is
e
W
il
c
oxon
te
s
t
to
c
om
pa
r
e
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
be
twe
e
n
di
f
f
e
r
e
nt
tr
e
a
tm
e
nt
s
of
th
e
s
a
m
e
a
lg
or
it
hm
s
.
T
he
B
e
nj
a
m
in
i
-
H
oc
hbe
r
g
(
B
H
)
m
e
th
od
is
u
s
e
d
f
or
p
-
va
lu
e
a
dj
us
tm
e
nt
.
T
he
r
e
s
ul
ts
in
T
a
bl
e
4
s
how
s
th
a
t
in
m
os
t
c
a
s
e
s
,
w
it
h
α=
0.05,
it
c
a
n
be
s
e
e
n
th
a
t
tr
e
a
tm
e
nt
s
f
or
dupl
ic
a
te
da
ta
s
ig
ni
f
ic
a
nt
ly
a
f
f
e
c
t
th
e
pe
r
f
or
m
a
n
c
e
.
T
he
R
2
dur
in
g
tr
a
in
in
g
w
it
h
H
G
B
R
of
th
e
unt
r
e
a
te
d
a
nd
a
ve
r
a
ge
d
tr
e
a
tm
e
nt
s
is
th
e
onl
y
c
om
pa
r
is
on
th
a
t
is
not
s
ig
ni
f
ic
a
nt
ly
di
f
f
e
r
e
nt
.
H
ow
e
ve
r
,
it
s
c
ount
e
r
pa
r
t
in
t
e
s
ti
ng i
s
s
ig
ni
f
ic
a
nt
ly
di
f
f
e
r
e
nt
.
T
a
bl
e
3
. R
e
s
ul
ts
of
t
he
r
e
pe
a
t
e
d m
e
a
s
ur
e
s
F
r
ie
dm
a
n t
e
s
t
of
t
he
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
be
twe
e
n t
r
e
a
tm
e
nt
s
A
l
gor
i
t
hm
M
e
t
r
i
c
s
n
F
D
e
gr
e
e
of
f
r
e
e
dom
p
-
va
l
ue
H
G
B
R
T
r
a
i
n R
2
12
18.667
2
<0.001
R
F
R
T
r
a
i
n R
2
12
22.167
2
<0.001
L
G
B
R
T
r
a
i
n R
2
12
22.167
2
<0.001
H
G
B
R
T
e
s
t
R
2
12
20.667
2
<0.001
R
F
R
T
e
s
t
R
2
12
22.167
2
<0.001
L
G
B
R
T
e
s
t
R
2
12
22.167
2
<0.001
H
G
B
R
T
r
a
i
n R
M
S
E
12
19.500
2
<0.001
R
F
R
T
r
a
i
n R
M
S
E
12
24.000
2
<0.001
L
G
B
R
T
r
a
i
n R
M
S
E
12
24.000
2
<0.001
H
G
B
R
T
e
s
t
R
M
S
E
12
24.000
2
<0.001
R
F
R
T
e
s
t
R
M
S
E
12
24.000
2
<0.001
L
G
B
R
T
e
s
t
R
M
S
E
12
24.000
2
<0.001
3.2
.
M
u
r
c
k
o f
r
ag
m
e
n
t
s
I
n
dr
ug
di
s
c
ove
r
y,
s
in
c
e
di
f
f
e
r
e
nt
f
r
a
gm
e
nt
s
le
a
d
to
di
f
f
e
r
e
nt
bi
oa
c
ti
vi
ty
be
twe
e
n
th
e
s
m
a
ll
m
ol
e
c
ul
e
s
a
nd
th
e
ta
r
ge
t,
de
c
om
pos
in
g
th
e
c
om
pound
s
in
to
f
r
a
gm
e
nt
s
is
a
c
om
m
on
ta
s
k
[
33]
.
T
he
M
ur
c
ko
f
r
a
gm
e
nt
s
,
pr
opos
e
d
by
B
e
m
is
a
nd
M
ur
c
ko
in
1996
[
34
]
,
is
a
w
id
e
ly
a
dopt
e
d
te
c
hni
que
,
in
c
lu
di
ng
in
M
L
D
D
[
35]
,
[
36]
.
T
he
m
e
th
od
w
or
ks
by
r
in
g
s
ys
te
m
s
,
li
nke
r
s
,
a
nd
t
he
s
id
e
c
ha
in
s
of
th
e
m
ol
e
c
ul
e
s
.
T
h
e
M
ur
c
ko
f
r
a
gm
e
nt
s
c
ons
is
t
of
a
c
om
bi
na
ti
on
of
r
in
gs
a
nd
li
nke
r
s
b
e
twe
e
n
th
e
m
,
w
it
h
a
ll
te
r
m
in
a
l
s
ubs
ti
tu
e
nt
s
r
e
m
ove
d.
I
n
th
is
pa
r
t,
w
e
c
om
pa
r
e
th
e
c
ha
r
a
c
te
r
is
ti
c
s
of
th
e
M
ur
c
ko
f
r
a
gm
e
nt
s
be
twe
e
n
tr
e
a
tm
e
nt
s
a
nd
m
ode
li
ng
s
ta
ge
s
to
id
e
nt
if
y
th
e
c
a
us
e
of
th
e
lo
w
-
pe
r
f
o
r
m
a
nc
e
m
e
tr
ic
s
e
ve
n
a
f
te
r
a
dopt
in
g
hype
r
pa
r
a
m
e
te
r
tu
ni
ng.
T
he
M
ur
c
ko
f
r
a
gm
e
nt
s
a
r
e
e
xt
r
a
c
t
e
d
f
r
om
th
e
c
om
po
unds
us
in
g
th
e
R
c
he
m
i
s
tr
y
de
ve
lo
pm
e
nt
ki
t
(
R
C
D
K
)
pa
c
ka
ge
ve
r
s
io
n
3.8.1
[
37]
.
T
he
m
in
im
um
f
r
a
gm
e
nt
s
iz
e
us
e
d
in
th
e
e
xt
r
a
c
ti
on
is
th
r
e
e
.
I
n
to
ta
l,
551
f
r
a
gm
e
nt
s
c
a
n
be
id
e
nt
if
ie
d
f
r
om
th
e
bi
oa
c
ti
vi
ty
da
ta
s
e
t.
T
he
f
r
a
gm
e
nt
s
a
r
e
num
be
r
e
d
f
r
om
F
001
to
F
551
a
c
c
or
di
ng
to
th
e
ir
f
r
e
que
nc
ie
s
in
th
e
da
ta
s
e
t.
O
ut
of
th
e
551
f
r
a
gm
e
nt
s
,
12
w
it
h
th
e
hi
ghe
s
t
f
r
e
que
nc
ie
s
w
e
r
e
s
e
le
c
te
d f
or
f
ur
th
e
r
a
na
ly
s
is
.
I
n
r
e
ga
r
ds
to
pI
C
50
a
s
th
e
r
e
gr
e
s
s
io
n
ta
r
ge
t
a
nd
th
e
na
tu
r
e
of
th
e
M
ur
c
ko
f
r
a
gm
e
nt
s
a
s
a
f
r
a
gm
e
nt
th
a
t
a
ppe
a
r
s
in
r
e
la
te
d
c
om
pound
s
,
w
hi
c
h
in
tu
r
n
a
f
f
e
c
ts
th
e
c
o
m
pounds
’
c
ha
r
a
c
te
r
is
ti
c
s
,
th
e
n
th
e
ir
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
ts
w
hi
c
h
a
r
e
us
e
d
a
s
f
e
a
tu
r
e
s
f
or
th
e
r
e
gr
e
s
s
io
n
a
lg
o
r
it
hm
s
,
im
pl
y
th
a
t
c
om
pounds
w
it
h
th
e
s
a
m
e
M
ur
c
ko
f
r
a
gm
e
nt
s
houl
d
ha
ve
s
im
il
a
r
pI
C
50
.
F
ig
u
r
e
3
s
how
s
th
e
di
s
t
r
ib
ut
io
ns
of
th
e
p
I
C
50
of
th
e
s
e
le
c
te
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
3003
-
3013
3008
M
ur
c
ko
f
r
a
gm
e
nt
s
f
or
tr
a
in
in
g
a
nd
te
s
ti
ng
in
a
ll
th
r
e
e
tr
e
a
tm
e
nt
s
.
F
r
om
th
e
12
s
a
m
pl
e
d
M
ur
c
ko
f
r
a
gm
e
nt
s
,
i
t
c
a
n be
s
e
e
n
f
r
om
F
ig
ur
e
3 s
om
e
f
r
a
gm
e
nt
s
ha
ve
di
f
f
e
r
e
nt
pI
C
50
di
s
tr
ib
ut
io
ns
, s
o
th
e
t
r
e
nd i
s
m
or
e
pr
onounc
e
d
w
he
n
th
e
dupl
ic
a
te
bi
oa
c
ti
vi
ty
da
ta
a
r
e
dr
oppe
d.
F
or
in
s
ta
nc
e
,
th
e
M
ur
c
ko
f
r
a
gm
e
nt
s
F
001,
F
002,
F
003,
a
nd
F
005
ha
ve
di
f
f
e
r
e
nt
p
I
C
50
di
s
tr
ib
ut
io
ns
.
S
t
il
l
in
th
e
dr
oppe
d
r
ow
,
s
in
c
e
it
ha
s
f
e
w
e
r
da
ta
,
th
e
r
e
a
r
e
c
a
s
e
s
w
he
r
e
c
e
r
ta
in
M
ur
c
ko
f
r
a
gm
e
nt
s
onl
y
e
xi
s
t
in
e
it
he
r
da
ta
s
e
t,
s
uc
h
a
s
h
a
ppe
ne
d
w
it
h
F
010
a
nd
F
011.
D
e
s
pi
te
th
e
M
ur
c
ko
f
r
a
gm
e
n
t
F
010
a
ls
o
onl
y
a
ppe
a
r
in
g
in
one
of
two
da
ta
s
e
ts
in
th
e
a
ve
r
a
ge
d
tr
e
a
tm
e
nt
,
it
c
a
n
be
s
e
e
n t
ha
t
th
e
boxplot
s
i
n t
he
r
e
s
pe
c
ti
ve
r
ow
ha
ve
s
im
il
a
r
pI
C
50
di
s
tr
ib
ut
io
ns
.
T
a
bl
e
4
. R
e
s
ul
ts
of
t
he
two
-
s
id
e
d P
a
ir
w
is
e
W
il
c
oxon te
s
t
w
it
h t
he
B
H
a
dj
us
tm
e
nt
on t
he
p
e
r
f
or
m
a
nc
e
m
e
tr
ic
s
be
twe
e
n t
r
e
a
tm
e
nt
s
A
l
gor
i
t
hm
M
e
t
r
i
c
s
T
r
e
a
t
m
e
nt
gr
oup
1
T
r
e
a
t
m
e
nt
gr
oup
2
n1
n2
W
p
-
va
l
ue
A
dj
us
t
e
d p
-
va
l
ue
H
G
B
R
T
r
a
i
n R
2
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
32
0.622
0.622
H
G
B
R
T
r
a
i
n R
2
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
78
<0.001
<0.001
H
G
B
R
T
r
a
i
n R
2
A
ve
r
a
ge
d
D
r
oppe
d
12
12
78
<0.001
<0.001
R
F
R
T
r
a
i
n R
2
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
1
0.001
0.001
R
F
R
T
r
a
i
n R
2
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
78
<0.001
<0.001
R
F
R
T
r
a
i
n R
2
A
ve
r
a
ge
d
D
r
oppe
d
12
12
78
<0.001
<0.001
L
G
B
R
T
r
a
i
n R
2
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
1
0.001
0.001
L
G
B
R
T
r
a
i
n R
2
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
78
<0.001
<0.001
L
G
B
R
T
r
a
i
n R
2
A
ve
r
a
ge
d
D
r
oppe
d
12
12
78
<0.001
<0.001
H
G
B
R
T
e
s
t
R
2
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
71
0.009
0.009
H
G
B
R
T
e
s
t
R
2
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
78
<0.001
<0.001
H
G
B
R
T
e
s
t
R
2
A
ve
r
a
ge
d
D
r
oppe
d
12
12
78
<0.001
<0.001
R
F
R
T
e
s
t
R
2
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
76
0.001
0.001
R
F
R
T
e
s
t
R
2
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
78
<0.001
<0.001
R
F
R
T
e
s
t
R
2
A
ve
r
a
ge
d
D
r
oppe
d
12
12
78
<0.001
<0.001
L
G
B
R
T
e
s
t
R
2
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
76
0.001
0.001
L
G
B
R
T
e
s
t
R
2
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
78
<0.001
<0.001
L
G
B
R
T
e
s
t
R
2
A
ve
r
a
ge
d
D
r
oppe
d
12
12
78
<0.001
<0.001
H
G
B
R
T
r
a
i
n R
M
S
E
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
67
0.027
0.027
H
G
B
R
T
r
a
i
n R
M
S
E
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
0
<0.001
0.001
H
G
B
R
T
r
a
i
n R
M
S
E
A
ve
r
a
ge
d
D
r
oppe
d
12
12
0
<0.001
0.001
R
F
R
T
r
a
i
n R
M
S
E
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
78
<0.001
<0.001
R
F
R
T
r
a
i
n R
M
S
E
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
0
<0.001
<0.001
R
F
R
T
r
a
i
n R
M
S
E
A
ve
r
a
ge
d
D
r
oppe
d
12
12
0
<0.001
<0.001
L
G
B
R
T
r
a
i
n R
M
S
E
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
78
<0.001
<0.001
L
G
B
R
T
r
a
i
n R
M
S
E
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
0
<0.001
<0.001
L
G
B
R
T
r
a
i
n R
M
S
E
A
ve
r
a
ge
d
D
r
oppe
d
12
12
0
<0.001
<0.001
H
G
B
R
T
e
s
t
R
M
S
E
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
0
<0.001
<0.001
H
G
B
R
T
e
s
t
R
M
S
E
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
0
<0.001
<0.001
H
G
B
R
T
e
s
t
R
M
S
E
A
ve
r
a
ge
d
D
r
oppe
d
12
12
0
<0.001
<0.001
R
F
R
T
e
s
t
R
M
S
E
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
0
<0.001
<0.001
R
F
R
T
e
s
t
R
M
S
E
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
0
<0.001
<0.001
R
F
R
T
e
s
t
R
M
S
E
A
ve
r
a
ge
d
D
r
oppe
d
12
12
0
<0.001
<0.001
L
G
B
R
T
e
s
t
R
M
S
E
U
nt
r
e
a
t
e
d
A
ve
r
a
ge
d
12
12
0
<0.001
<0.001
L
G
B
R
T
e
s
t
R
M
S
E
U
nt
r
e
a
t
e
d
D
r
oppe
d
12
12
0
<0.001
<0.001
L
G
B
R
T
e
s
t
R
M
S
E
A
ve
r
a
ge
d
D
r
oppe
d
12
12
0
<0.001
<0.001
F
ig
ur
e
3. T
he
boxplot
s
of
t
he
pI
C
50
di
s
tr
ib
ut
io
ns
f
or
t
he
s
e
le
c
te
d M
ur
c
ko
f
r
a
gm
e
nt
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
I
nv
e
s
ti
gat
io
n on low
-
pe
r
fo
r
m
anc
e
t
une
d
-
r
e
g
r
e
s
s
or
of
i
nhi
bi
to
r
y
c
onc
e
nt
r
at
io
n …
(
D
ani
e
l
F
e
br
ia
n Se
ng
k
e
y
)
3009
F
ig
ur
e
4
s
how
s
th
e
12
s
e
le
c
te
d
M
ur
c
ko
f
r
a
gm
e
nt
s
pl
ot
te
d
a
s
l
in
e
s
tr
uc
tu
r
e
,
f
ol
lo
w
e
d
by
th
e
na
m
e
,
a
nd
s
ta
ti
s
ti
c
s
f
or
e
a
c
h
tr
e
a
tm
e
nt
.
A
s
th
e
s
pl
it
ti
ng
s
tr
a
te
gy
us
e
d
a
n
80:
20
pr
opo
r
ti
on
f
o
r
t
r
a
in
in
g
a
nd
te
s
ti
ng,
r
e
s
pe
c
ti
ve
ly
,
it
c
a
n
be
s
e
e
n
th
a
t
not
a
ll
of
th
e
s
e
s
e
le
c
te
d
f
r
a
gm
e
nt
s
a
r
e
e
ve
nl
y
di
s
tr
ib
ut
e
d
r
e
ga
r
di
ng
th
e
pr
opor
ti
on.
F
o
r
in
s
ta
nc
e
,
in
e
a
c
h
tr
e
a
tm
e
nt
,
th
e
r
e
30
c
om
pounds
s
ha
r
e
M
ur
c
ko
f
r
a
gm
e
nt
F
001.
I
n
th
e
unt
r
e
a
te
d
dupl
ic
a
te
s
da
ta
s
e
t,
th
e
s
pl
it
is
e
xa
c
tl
y
80:
20
(
24:
6)
,
b
ut
in
th
e
a
ve
r
a
ge
d
a
nd
dr
oppe
d,
th
e
s
pl
it
s
a
r
e
s
li
ght
ly
s
hi
f
te
d
to
86.67:13.33
(
26:
4)
.
F
002
is
a
not
he
r
f
r
e
qu
e
nt
M
ur
c
ko
f
r
a
gm
e
nt
,
th
a
t
s
pl
it
w
it
h
a
r
a
ti
o
78.94:21.06
(
30:
8)
,
80:
20
(
20:
5)
,
a
nd
90:
10
(
18:
2
)
a
t
th
e
unt
r
e
a
te
d,
a
ve
r
a
ge
d,
a
nd
dr
oppe
d
dupl
ic
a
te
tr
e
a
tm
e
nt
s
,
r
e
s
pe
c
ti
ve
ly
.
T
he
r
a
ti
o
f
or
th
e
M
ur
c
ko
f
r
a
gm
e
nt
F
002
a
t
th
e
d
r
oppe
d
tr
e
a
tm
e
nt
ha
s
a
m
a
jo
r
de
vi
a
ti
on
f
r
om
th
e
e
xpe
c
te
d
s
pl
it
r
a
ti
o.
T
h
e
de
vi
a
ti
ons
of
th
e
s
pl
it
r
a
ti
o
a
r
e
e
ve
n
m
or
e
not
ic
e
a
bl
e
f
or
th
e
s
e
le
c
te
d
M
ur
c
ko
f
r
a
gm
e
nt
s
w
it
h
le
s
s
f
r
e
que
nc
y,
s
uc
h
a
s
F
010
a
nd
F
011.
M
ur
c
ko
f
r
a
gm
e
nt
F
010
w
a
s
di
s
tr
ib
ut
e
d w
it
h a
r
a
ti
o of
75:
5 (
21:
7)
f
or
t
he
unt
r
e
a
te
d
dupl
ic
a
t
e
a
nd 100:0 f
or
t
he
ot
he
r
t
w
o t
r
e
a
tm
e
nt
s
.
F
ig
ur
e
4. S
e
le
c
te
d M
ur
c
ko f
r
a
gm
e
nt
s
T
he
f
ir
s
t
li
ne
in
e
a
c
h
c
e
ll
s
how
s
th
e
f
r
a
gm
e
nt
num
be
r
(
F
###)
.
T
he
n
th
e
s
e
c
ond,
th
ir
d,
a
nd
f
our
th
li
ne
s
s
how
t
he
pr
opor
ti
on a
nd p
I
C
50
s
ta
ti
s
ti
c
s
i
n untr
e
a
te
d dupli
c
a
te
s
, a
ve
r
a
ge
d pI
C
50
, a
nd dr
oppe
d dupli
c
a
te
s
,
r
e
s
pe
c
ti
ve
ly
.
I
n
e
a
c
h
li
ne
,
th
e
num
be
r
s
s
ho
w
f
r
e
que
nc
ie
s
a
nd
pr
opor
ti
ons
of
th
e
r
e
s
pe
c
ti
ve
f
r
a
gm
e
nt
in
th
e
tr
a
in
in
g/
te
s
ti
ng
da
ta
s
e
t,
f
ol
lo
w
e
d
by
th
e
r
e
s
pe
c
ti
ve
a
ve
r
a
g
e
a
nd
s
ta
nd
a
r
d
de
vi
a
ti
on
of
pI
C
50
in
th
e
tr
a
in
in
g/
te
s
ti
ng da
ta
s
e
t
.
3.3
.
D
is
c
u
s
s
io
n
s
T
ypi
c
a
ll
y,
hype
r
pa
r
a
m
e
te
r
tu
ni
ng
is
a
ppl
ie
d
to
ga
in
hi
ghe
r
M
L
m
ode
l
pe
r
f
or
m
a
nc
e
s
uc
h
a
s
de
m
ons
tr
a
te
d
in
pr
e
vi
ous
s
tu
di
e
s
[
38]
.
H
ow
e
ve
r
,
e
ve
n
w
it
h
a
la
r
ge
hype
r
pa
r
a
m
e
te
r
s
s
e
a
r
c
h
s
pa
c
e
,
in
th
is
pa
r
ti
c
ul
a
r
s
tu
dy,
w
e
f
ound
th
e
r
e
gr
e
s
s
or
s
’
pe
r
f
or
m
a
nc
e
s
w
e
r
e
not
a
s
e
xpe
c
te
d.
T
he
r
e
f
or
e
,
by
c
onduc
ti
ng
f
ur
th
e
r
a
na
ly
s
e
s
,
w
e
a
ppl
ie
d s
ta
ti
s
ti
c
a
l
t
e
s
ts
to
th
e
pe
r
f
or
m
a
nc
e
da
ta
,
gr
oupe
d
by
th
e
tr
e
a
tm
e
nt
s
f
or
dupl
ic
a
te
d
bi
oa
c
ti
vi
ty
da
ta
.
T
he
r
e
s
ul
ts
of
th
e
r
e
pe
a
te
d
m
e
a
s
ur
e
s
F
r
ie
dm
a
n
te
s
t
s
how
th
a
t
th
e
di
f
f
e
r
e
nc
e
s
in
th
e
da
ta
pr
e
pa
r
a
ti
on
s
ig
ni
f
ic
a
nt
ly
im
pa
c
t
m
ode
l
pe
r
f
or
m
a
nc
e
,
r
e
ga
r
dl
e
s
s
of
th
e
a
lg
or
it
hm
s
.
T
hi
s
f
in
di
ng
is
c
on
s
is
te
nt
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
3003
-
3013
3010
w
it
h
pr
e
vi
ous
s
tu
di
e
s
on
hype
r
pa
r
a
m
e
te
r
opt
im
iz
a
ti
on
.
A
s
tu
dy
by
S
c
hr
a
tz
e
t
al
.
[
39]
on
hype
r
pa
r
a
m
e
te
r
tu
ni
ng
in
th
e
f
ie
ld
of
e
c
ol
ogi
c
a
l
m
ode
li
ng,
it
w
a
s
f
ound
th
a
t
t
he
r
e
s
ul
ts
of
hype
r
pa
r
a
m
e
te
r
tu
ni
ng
m
ig
ht
be
ne
gl
ig
ib
le
f
or
R
F
.
S
im
il
a
r
ly
,
S
ip
pe
r
[
40
]
e
va
lu
a
te
d
m
a
ny
a
lg
or
i
th
m
s
a
nd
da
ta
s
e
ts
a
nd
f
ound
th
a
t
c
on
s
id
e
r
a
bl
e
ga
in
s
c
oul
d
not
a
lw
a
ys
be
e
xpe
c
t
e
d
f
r
om
hype
r
pa
r
a
m
e
te
r
tu
ni
ng.
T
he
s
tu
dy
a
ls
o
f
ound
th
a
t
R
F
R
,
w
hi
c
h
w
a
s
a
ls
o us
e
d i
n our
s
tu
dy, i
s
on
e
a
lg
or
it
hm
e
xpe
c
te
d t
o ga
in
l
e
s
s
f
r
om
hype
r
pa
r
a
m
e
te
r
t
uni
ng.
S
pl
it
ti
ng
th
e
da
ta
s
e
t
f
or
tr
a
in
in
g
a
nd
te
s
ti
ng
is
a
s
ta
nda
r
d
pr
a
c
ti
c
e
in
M
L
.
I
n
c
la
s
s
if
ic
a
ti
on
ta
s
k
s
,
e
ns
ur
in
g
th
e
ba
la
nc
e
be
twe
e
n
th
e
la
be
ls
or
c
la
s
s
e
s
is
a
n
im
por
t
a
nt
c
ons
id
e
r
a
ti
on
in
da
ta
pr
e
pa
r
a
ti
on
s
in
c
e
th
e
di
ve
r
s
it
y of
t
he
s
a
m
pl
e
s
i
n e
a
c
h c
la
s
s
br
in
gs
c
on
s
id
e
r
a
bl
e
i
nf
lu
e
nc
e
t
o t
he
m
ode
l
pe
r
f
or
m
a
nc
e
[
41]
. I
n a
not
he
r
s
tu
dy
of
he
a
r
t
di
s
e
a
s
e
c
la
s
s
if
ic
a
ti
on
w
it
h
e
ns
e
m
bl
e
a
lg
or
it
hm
s
,
th
e
pr
e
s
e
r
ve
d
di
s
tr
ib
ut
io
n
in
tr
a
in
-
te
s
t
s
pl
it
ti
ng
br
ought
c
ons
id
e
r
a
bl
e
im
pa
c
ts
to
th
e
ove
r
a
ll
pe
r
f
or
m
a
nc
e
[
42]
.
P
r
e
di
c
ti
on
ta
s
ks
s
uc
h
a
s
r
e
gr
e
s
s
io
ns
do
not
s
ha
r
e
th
is
da
ta
s
e
t
im
ba
la
nc
e
pr
obl
e
m
due
to
th
e
d
if
f
e
r
e
nt
na
tu
r
e
of
th
e
ta
r
ge
t.
H
ow
e
ve
r
,
th
e
r
e
pr
e
s
e
nt
a
ti
ve
ne
s
s
of
th
e
d
a
ta
c
ha
r
a
c
te
r
is
ti
c
s
di
s
tr
ib
ut
io
n
in
bot
h
tr
a
in
in
g
a
nd
te
s
ti
ng
da
ta
s
e
ts
ha
s
to
be
c
ons
id
e
r
e
d
.
T
hi
s
im
pl
ie
s
th
a
t
th
e
f
a
ir
ne
s
s
of
da
ta
c
ha
r
a
c
te
r
is
ti
c
s
in
th
e
tr
a
in
-
te
s
t
s
pl
it
ha
s
to
be
c
ons
id
e
r
e
d,
a
s
pr
opos
e
d
in
th
e
s
tu
dy
by
S
a
l
a
z
a
r
e
t
al
.
[
43]
.
I
n
th
is
s
tu
dy,
r
e
g
a
r
dl
e
s
s
of
th
e
hype
r
pa
r
a
m
e
te
r
tu
ni
ng
w
it
h
a
n
e
xha
us
ti
ve
s
e
a
r
c
h
s
p
a
c
e
on
v
a
r
io
us
c
om
bi
na
ti
ons
of
tr
e
a
tm
e
nt
s
of
dupl
ic
a
te
s
,
f
e
a
tu
r
e
e
xt
r
a
c
ti
on
us
in
g
v
a
r
io
us
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
ts
a
s
de
s
c
r
ip
to
r
s
,
a
nd
s
e
ve
r
a
l
a
lg
or
it
hm
s
,
t
he
be
s
t
m
ode
ls
s
ti
ll
ha
ve
lo
w
pe
r
f
or
m
a
nc
e
.
A
s
th
e
M
ur
c
ko
f
r
a
gm
e
nt
r
e
pr
e
s
e
nt
s
th
e
c
or
e
s
tr
uc
tu
r
a
l
f
r
a
m
e
w
or
k
of
a
m
ol
e
c
ul
e
,
in
c
lu
di
ng
it
s
r
in
gs
a
nd
li
nke
r
s
,
w
it
h
th
e
s
id
e
c
ha
in
s
or
t
e
r
m
in
a
l
s
ubs
ti
tu
e
nt
s
e
xc
lu
de
d
,
it
is
c
e
nt
r
a
l
to
th
e
m
ol
e
c
ul
a
r
s
tr
uc
tu
r
e
a
nd
of
te
n
c
ons
id
e
r
e
d
a
s
th
e
s
c
a
f
f
ol
d
on
w
hi
c
h
va
r
io
us
f
unc
ti
ona
l
gr
oups
a
r
e
a
tt
a
c
he
d.
O
ur
in
ve
s
ti
ga
ti
on
of
th
e
M
ur
c
ko
f
r
a
gm
e
nt
s
di
s
tr
ib
ut
io
ns
in
th
e
tr
a
in
a
nd
te
s
t
da
ta
s
e
t
s
f
ound
th
a
t
s
om
e
of
th
e
m
w
e
r
e
not
e
qua
ll
y
di
s
tr
ib
ut
e
d
in
bot
h
da
ta
s
e
ts
,
r
e
s
ul
ti
ng
in
a
f
r
a
gm
e
nt
s
im
ba
la
nc
e
b
e
twe
e
n
th
e
da
ta
s
e
ts
,
th
e
r
e
f
or
e
,
th
e
f
e
a
tu
r
e
s
le
a
r
ne
d
by
th
e
m
ode
ls
a
r
e
di
f
f
e
r
e
nt
f
r
om
th
os
e
in
th
e
te
s
t
da
ta
s
e
t.
T
hi
s
is
s
ue
s
houl
d
be
c
on
s
id
e
r
e
d
f
ur
th
e
r
w
it
h
a
n
e
xpa
nde
d
li
s
t
of
a
lg
or
it
hm
s
a
nd bioa
c
ti
vi
ty
t
a
r
ge
ts
.
4.
C
O
N
C
L
U
S
I
O
N
I
n
th
is
s
tu
dy,
w
e
in
ve
s
ti
ga
te
d
th
e
lo
w
pe
r
f
or
m
a
nc
e
of
th
e
e
ns
e
m
bl
e
tr
e
e
-
ba
s
e
d
r
e
gr
e
s
s
or
a
lg
or
it
hm
s
in
pr
e
di
c
ti
ng
th
e
I
C
50
of
s
m
a
ll
m
ol
e
c
ul
e
s
,
ta
r
ge
ti
ng
th
e
S
A
R
S
-
C
oV
-
2
pp1a
b.
D
e
s
pi
te
th
e
e
xha
u
s
ti
ve
hype
r
pa
r
a
m
e
te
r
s
e
a
r
c
h
s
pa
c
e
,
va
r
io
us
c
om
bi
na
ti
ons
of
tr
e
a
tm
e
nt
s
of
dupl
ic
a
te
bi
oa
c
ti
vi
ty
da
ta
a
nd
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
t
de
s
c
r
ip
to
r
s
a
s
f
e
a
tu
r
e
s
,
none
of
th
e
r
e
s
ul
ti
ng
m
o
de
ls
ga
in
e
d
a
s
a
ti
s
f
a
c
to
r
y
num
b
e
r
of
R
2
a
nd
R
M
S
E
. T
r
e
a
tm
e
nt
-
w
is
e
, dr
oppi
ng a
ll
t
he
dupli
c
a
te
d bi
oa
c
ti
vi
ty
da
ta
yi
e
ld
e
d t
he
w
or
s
t
pe
r
f
or
m
a
nc
e
c
om
pa
r
e
d
to
th
e
ot
he
r
two
tr
e
a
tm
e
nt
s
.
T
he
R
2
v
a
lu
e
s
a
c
r
os
s
m
ode
li
ng
s
t
a
ge
s
(
tr
a
in
,
c
r
o
s
s
-
va
li
da
ti
on,
a
nd
te
s
t)
te
nd
to
ha
ve
s
im
il
a
r
t
r
e
nds
r
e
ga
r
dl
e
s
s
of
th
e
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
ts
a
nd a
lg
or
it
hm
s
. H
ow
e
ve
r
, a
de
e
pe
r
c
om
pa
r
is
on of
th
e
R
M
S
E
in
e
a
c
h
m
ol
e
c
ul
a
r
f
in
ge
r
pr
in
t
s
how
s
th
a
t
th
e
e
xpe
r
i
m
e
nt
s
w
it
h
unt
r
e
a
te
d
dupl
ic
a
te
s
te
nd
to
yi
e
ld
hi
ghe
r
R
M
S
E
in
te
s
t
c
r
os
s
-
va
li
da
ti
on
th
a
n
in
th
e
r
e
a
l
tr
a
in
in
g
d
a
ta
s
e
t.
A
t
th
e
s
a
m
e
ti
m
e
,
a
s
a
lo
s
s
f
unc
ti
on,
it
s
houl
d
be
th
e
ot
he
r
w
a
y
a
r
ound.
H
e
nc
e
,
ba
s
e
d
on
our
e
xpe
r
i
m
e
nt
s
,
tr
e
a
ti
ng
th
e
dupl
ic
a
te
s
by
a
ve
r
a
gi
ng
th
e
pI
C
50
br
ought
m
or
e
r
e
a
s
ona
bl
e
r
e
s
ul
t
s
.
T
he
ba
la
nc
e
d
di
s
tr
ib
ut
io
n
be
twe
e
n
la
be
ls
i
s
a
n
im
por
ta
nt
f
a
c
to
r
in
ove
r
a
ll
m
ode
l
pe
r
f
or
m
a
nc
e
in
c
la
s
s
if
ic
a
ti
on
ta
s
k
s
.
B
y
ha
vi
ng
b
a
la
nc
e
d
la
be
l
di
s
tr
ib
ut
io
n
in
bot
h
tr
a
in
in
g
a
nd
te
s
ti
ng
da
ta
s
e
ts
,
th
e
c
ons
i
s
te
nc
y
of
th
e
d
a
ta
c
oul
d
b
e
pr
e
s
e
r
ve
d,
he
nc
e
,
th
e
c
ha
r
a
c
t
e
r
is
ti
c
s
f
a
c
e
d
by
th
e
a
lg
or
it
hm
dur
in
g
m
ode
l
tr
a
in
in
g
c
oul
d
a
ls
o
be
f
ound
w
he
n
e
va
lu
a
ti
ng
th
e
m
ode
l
w
it
h
th
e
te
s
ti
ng
da
ta
s
e
t.
R
e
ga
r
dl
e
s
s
of
th
e
na
tu
r
e
of
th
e
ta
s
k,
th
e
r
e
pr
e
s
e
nt
a
ti
ve
ne
s
s
of
th
e
c
ha
r
a
c
te
r
is
ti
c
s
in
th
e
tr
a
in
in
g
a
nd
te
s
ti
ng
da
ta
s
e
ts
a
l
s
o
in
f
lu
e
nc
e
s
th
e
m
ode
l
pe
r
f
or
m
a
nc
e
.
I
n
ou
r
s
tu
dy
,
our
in
ve
s
ti
ga
ti
on
of
th
e
M
ur
c
ko
f
r
a
gm
e
nt
s
di
s
tr
ib
ut
io
ns
i
n t
he
da
ta
s
e
ts
us
e
d f
or
t
r
a
in
in
g a
nd t
e
s
ti
ng w
a
s
no
t
ba
la
nc
e
d. T
he
r
e
a
r
e
c
a
s
e
s
w
he
r
e
s
om
e
of
t
he
f
r
e
que
nt
M
ur
c
ko
f
r
a
gm
e
nt
s
in
th
e
w
hol
e
da
ta
s
e
t
w
e
r
e
not
e
ve
nl
y
di
s
tr
ib
ut
e
d
or
di
d
not
e
xi
s
t
in
th
e
te
s
ti
ng
da
ta
s
e
t.
T
hi
s
is
c
ons
id
e
r
e
d
th
e
m
a
in
c
a
u
s
e
of
th
e
m
ode
ls
,
de
s
pi
te
hype
r
pa
r
a
m
e
te
r
s
be
in
g
tu
ne
d
w
it
h
a
n
e
xha
us
ti
ve
li
s
t
of
s
e
a
r
c
h
s
p
a
c
e
,
w
hi
c
h
te
nds
to
ove
r
f
it
.
F
ut
ur
e
s
tu
di
e
s
s
houl
d
c
ons
id
e
r
th
e
is
s
ue
of
M
ur
c
ko
f
r
a
gm
e
nt
di
s
tr
ib
ut
io
n.
W
he
n
in
ve
s
ti
ga
ti
ng
th
e
e
f
f
e
c
t
of
M
ur
c
ko
f
r
a
gm
e
nt
di
s
tr
ib
ut
io
ns
in
qua
nt
it
a
ti
ve
s
tr
uc
tu
r
e
-
a
c
ti
vi
ty
r
e
la
ti
ons
hi
p
(
Q
S
A
R
)
m
ode
li
ng,
a
w
id
e
r
a
ng
e
of
a
lg
or
it
hm
s
,
ta
r
ge
ts
,
ta
s
k
s
,
a
nd
s
pl
it
r
a
ti
os
m
us
t
be
c
ons
id
e
r
e
d.
A
C
K
N
O
WL
E
D
G
E
M
E
N
T
S
T
he
a
ut
hor
s
th
a
nk
th
e
s
ta
f
f
of
th
e
I
nf
or
m
a
ti
on
a
nd
C
om
m
uni
c
a
ti
on
T
e
c
hnol
ogy
A
c
a
de
m
ic
S
uppor
t
U
ni
t
, U
ni
ve
r
s
it
a
s
S
a
m
R
a
tu
la
ngi
, f
or
t
e
c
hni
c
a
l
s
uppor
t
dur
in
g t
he
e
xpe
r
im
e
nt
s
.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
T
hi
s
w
or
k
is
f
unde
d
by
th
e
D
a
f
ta
r
I
s
ia
n
P
e
la
ks
a
na
a
n
A
ngga
r
a
n
(
D
I
P
A
)
U
ni
ve
r
s
it
a
s
S
a
m
R
a
tu
la
ngi
:
R
is
e
t
D
a
s
a
r
U
nggula
n U
N
S
R
A
T
2024,
c
ont
r
a
c
t
num
be
r
:
184/
U
N
12.13/L
T
/2
024.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
I
nv
e
s
ti
gat
io
n on low
-
pe
r
fo
r
m
anc
e
t
une
d
-
r
e
g
r
e
s
s
or
of
i
nhi
bi
to
r
y
c
onc
e
nt
r
at
io
n …
(
D
ani
e
l
F
e
br
ia
n Se
ng
k
e
y
)
3011
A
U
T
H
O
R
C
O
N
T
R
I
B
U
T
I
O
N
S
S
T
A
T
E
M
E
N
T
T
hi
s
jo
ur
na
l
us
e
s
th
e
C
ont
r
ib
ut
or
R
ol
e
s
T
a
xonomy
(
C
R
e
di
T
)
to
r
e
c
ogni
z
e
in
di
vi
dua
l
a
ut
hor
c
ont
r
ib
ut
io
ns
, r
e
duc
e
a
ut
hor
s
hi
p di
s
put
e
s
,
a
nd f
a
c
il
it
a
te
c
ol
la
bo
r
a
ti
on.
N
am
e
o
f
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
D
a
ni
e
l
F
e
br
ia
n
S
e
ngke
y
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
A
nge
li
na
S
te
va
ny
R
e
gi
na
M
a
s
e
ngi
✓
✓
✓
✓
✓
✓
✓
✓
✓
A
lwi
n M
e
lk
ie
S
a
m
bul
✓
✓
✓
✓
✓
✓
✓
✓
T
r
in
a
E
ka
w
a
ti
T
a
ll
e
i
✓
✓
✓
✓
✓
S
he
r
w
in
R
e
in
a
ld
o
U
ns
r
a
td
ia
nt
o S
om
pi
e
✓
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
l
i
z
a
t
i
on
M
:
M
e
t
hodol
ogy
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
da
t
i
on
Fo
:
Fo
r
m
a
l
a
na
l
ys
i
s
I
:
I
nve
s
t
i
ga
t
i
on
R
:
R
e
s
our
c
e
s
D
:
D
a
t
a
C
ur
a
t
i
on
O
:
W
r
i
t
i
ng
-
O
r
i
gi
na
l
D
r
a
f
t
E
:
W
r
i
t
i
ng
-
R
e
vi
e
w
&
E
di
t
i
ng
Vi
:
Vi
s
ua
l
i
z
a
t
i
on
Su
:
Su
pe
r
vi
s
i
on
P
:
P
r
oj
e
c
t
a
dm
i
ni
s
t
r
a
t
i
on
Fu
:
Fu
ndi
ng a
c
qui
s
i
t
i
on
C
O
N
F
L
I
C
T
O
F
I
N
T
E
R
E
S
T
S
T
A
T
E
M
E
N
T
T
he
a
ut
hor
s
s
t
a
te
no c
onf
li
c
t
of
i
nt
e
r
e
s
t.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
T
he
bi
oa
c
ti
vi
ty
da
ta
s
e
t
us
e
d
in
th
is
r
e
s
e
a
r
c
h
w
a
s
r
e
tr
ie
ve
d
f
r
om
th
e
C
hE
M
B
L
da
ta
b
a
s
e
a
t
ht
tp
s
:/
/e
bi
.a
c
.uk/
c
he
m
bl
in
A
pr
il
2024.
P
r
e
pr
oc
e
s
s
e
d
da
ta
s
e
ts
w
it
h
e
xt
r
a
c
te
d
f
in
ge
r
pr
in
ts
a
s
f
e
a
tu
r
e
s
a
nd
th
e
di
f
f
e
r
e
nt
tr
e
a
tm
e
nt
s
on
th
e
dupl
ic
a
te
bi
oa
c
ti
vi
ty
a
r
e
a
va
il
a
bl
e
in
ht
tp
s
:/
/g
it
hub.c
om
/d
a
ni
e
ls
e
ngke
y/
s
uppl
e
m
e
nt
a
r
ie
s
/t
r
e
e
/m
a
in
/p
p
1a
b_i
ja
i2
025.
R
E
F
E
R
E
N
C
E
S
[
1]
X
.
H
ua
ng,
R
.
P
e
a
r
c
e
,
G
.
S
.
O
m
e
nn,
a
nd
Y
. Z
ha
ng,
“
I
de
nt
i
f
i
c
a
t
i
on
o
f
13
gua
ni
d
i
nobe
nz
oyl
-
or
a
m
i
ni
di
nobe
nz
oyl
-
c
ont
a
i
ni
ng
dr
ugs
t
o
pot
e
nt
i
a
l
l
y
i
nhi
bi
t
T
M
P
R
S
S
2
f
or
C
O
V
I
D
-
19
t
r
e
a
t
m
e
nt
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
M
ol
e
c
ul
ar
Sc
i
e
nc
e
s
,
vol
.
22,
no.
13,
J
un. 2021, doi
:
10.3390/
i
j
m
s
22137060.
[
2]
R
.
A
l
e
xpa
ndi
,
J
.
F
.
D
e
M
e
s
qui
t
a
,
S
.
K
.
P
a
ndi
a
n,
a
nd
A
.
V
.
R
a
vi
,
“
Q
ui
nol
i
ne
s
-
ba
s
e
d
S
A
R
S
-
C
oV
-
2
3C
L
pr
o
a
nd
R
dR
p
i
nhi
bi
t
or
s
a
nd
s
pi
ke
-
R
B
D
-
A
C
E
2
i
nhi
bi
t
or
f
or
dr
ug
-
r
e
pur
pos
i
ng
a
ga
i
ns
t
C
O
V
I
D
-
19:
a
n
i
n
s
i
l
i
c
o
a
na
l
ys
i
s
,”
F
r
ont
i
e
r
s
i
n
M
i
c
r
obi
ol
ogy
,
vol
. 11, J
ul
. 2020, doi
:
10.3389/
f
m
i
c
b.2020.01796.
[
3]
A
.
K
ha
l
e
d
a
nd
Z
.
A
.
E
l
H
a
l
i
e
m
,
“
G
e
ne
r
a
t
i
ve
r
e
c
ur
r
e
nt
ne
t
w
or
k
f
or
de
s
i
gn
S
A
R
S
-
C
oV
-
2
m
a
i
n
p
r
ot
e
a
s
e
i
nhi
bi
t
or
,”
i
n
2022
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
Sof
t
w
ar
e
,
T
e
l
e
c
om
m
uni
c
at
i
ons
and
C
om
put
e
r
N
e
t
w
o
r
k
s
(
Sof
t
C
O
M
)
,
2022,
pp.
1
–
6
,
doi
:
10.23919/
S
of
t
C
O
M
55329.2022.9911377.
[
4]
D
.
S
ha
j
i
,
S
.
Y
a
m
a
m
ot
o,
R
.
S
a
i
t
o,
R
.
S
uz
uki
,
S
.
N
a
ka
m
ur
a
,
a
nd
N
.
K
u
r
i
t
a
,
“
P
r
opos
a
l
of
nove
l
na
t
ur
a
l
i
nhi
bi
t
o
r
s
of
s
e
ve
r
e
a
c
ut
e
r
e
s
pi
r
a
t
or
y
s
yndr
om
e
c
or
ona
vi
r
us
2
m
a
i
n
pr
ot
e
a
s
e
:
m
ol
e
c
ul
a
r
doc
ki
ng
a
nd
a
b
i
ni
t
i
o
f
r
a
gm
e
nt
m
ol
e
c
ul
a
r
or
bi
t
a
l
c
a
l
c
ul
a
t
i
ons
,”
B
i
ophy
s
i
c
al
C
he
m
i
s
t
r
y
, vol
. 275, A
ug. 2021, doi
:
10.1016/
j
.bpc
.2021.106608.
[
5]
F
.
H
u,
D
.
W
a
ng,
Y
.
H
u,
J
.
J
i
a
ng,
a
nd
P
.
Y
i
n,
“
G
e
ne
r
a
t
i
ng
nove
l
c
om
pound
s
t
a
r
ge
t
i
ng
S
A
R
S
-
C
oV
-
2
m
a
i
n
pr
ot
e
a
s
e
ba
s
e
d
on
i
m
ba
l
a
nc
e
d
da
t
a
s
e
t
,”
i
n
2020
I
E
E
E
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
B
i
oi
nf
or
m
at
i
c
s
and
B
i
om
e
di
c
i
ne
(
B
I
B
M
)
,
I
E
E
E
,
D
e
c
.
2020,
pp. 432
–
436
, doi
:
10.1109/
B
I
B
M
49941.2020.9313317.
[
6]
I
.
A
c
hi
l
onu,
E
.
A
.
I
w
uc
hukw
u,
O
.
J
.
A
c
hi
l
onu,
M
.
A
.
F
e
r
na
nde
s
,
a
nd
Y
.
S
a
ye
d
,
“
T
a
r
ge
t
i
ng
t
he
S
A
R
S
-
C
oV
-
2
m
a
i
n
p
r
ot
e
a
s
e
us
i
ng
F
D
A
-
a
ppr
ove
d
i
s
a
vuc
ona
z
oni
um
,
a
P
2
–
P
3
α
-
ke
t
oa
m
i
de
de
r
i
va
t
i
ve
a
nd
pe
nt
a
ga
s
t
r
i
n:
a
n
in
-
s
i
l
i
c
o
dr
ug
di
s
c
ove
r
y
a
ppr
oa
c
h,”
J
our
nal
of
M
ol
e
c
ul
ar
G
r
aphi
c
s
and M
ode
l
l
i
ng
, vol
. 101, D
e
c
. 2020, doi
:
10.10
16/
j
.j
m
gm
.2020.107730.
[
7]
N
.
F
e
r
dous
e
t
al
.
,
“
M
pr
opr
e
d:
a
m
a
c
hi
ne
l
e
a
r
ni
ng
(
M
L
)
dr
i
ve
n
w
e
b
-
a
pp
f
or
bi
oa
c
t
i
vi
t
y
pr
e
di
c
t
i
on
of
S
A
R
S
-
C
oV
-
2
m
a
i
n
pr
ot
e
a
s
e
(M
pr
o
)
a
nt
a
goni
s
t
s
,”
P
L
O
S O
N
E
, vol
. 18, no. 6, J
un. 2023, doi
:
10.1371/
j
our
na
l
.
pone
.0287179.
[
8]
T
.
E
.
T
a
l
l
e
i
e
t
al
.
,
“
P
ot
e
nt
i
a
l
of
pl
a
nt
bi
oa
c
t
i
ve
c
om
pounds
a
s
S
A
R
S
-
C
oV
-
2
m
a
i
n
pr
ot
e
a
s
e
(
M
pr
o
)
a
nd
s
pi
ke
(
S
)
gl
yc
opr
ot
e
i
n
i
nhi
bi
t
or
s
:
a
m
ol
e
c
ul
a
r
doc
ki
ng s
t
udy,”
Sc
i
e
nt
i
f
i
c
a
, vol
. 2020, pp. 1
–
18, D
e
c
. 2
020, doi
:
10.1155/
2020/
6307457.
[
9]
T
.
E
.
T
a
l
l
e
i
e
t
al
.
,
“
F
r
ui
t
b
r
om
e
l
a
i
n
-
de
r
i
ve
d
pe
pt
i
de
pot
e
nt
i
a
l
l
y
r
e
s
t
r
a
i
ns
t
he
a
t
t
a
c
hm
e
nt
of
S
A
R
S
-
C
oV
-
2
va
r
i
a
nt
s
t
o
hA
C
E
2:
a
pha
r
m
a
c
oi
nf
or
m
a
t
i
c
s
a
ppr
oa
c
h,”
M
ol
e
c
ul
e
s
, vol
. 27, no. 1, J
a
n. 2022, doi
:
10.3390/
m
ol
e
c
ul
e
s
27010260.
[
10]
F
.
S
ul
i
s
t
i
a
w
a
n,
W
.
A
.
K
us
um
a
,
N
.
S
.
R
a
m
a
dha
nt
i
,
a
nd
A
.
T
e
dj
o,
“
D
r
ug
-
t
a
r
ge
t
i
nt
e
r
a
c
t
i
on
pr
e
di
c
t
i
on
i
n
c
or
ona
vi
r
us
d
i
s
e
a
s
e
2019
c
a
s
e
us
i
ng
de
e
p
s
e
m
i
-
s
upe
r
vi
s
e
d
l
e
a
r
ni
ng
m
ode
l
,”
i
n
2020
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
A
dv
anc
e
d
C
om
put
e
r
Sc
i
e
nc
e
and
I
nf
or
m
at
i
on Sy
s
t
e
m
s
, I
E
E
E
, O
c
t
. 2020, pp. 83
–
88
, doi
:
10.1109/
I
C
A
C
S
I
S
51025
.2020.9263241.
[
11]
L
.
E
r
l
i
na
e
t
al
.
,
“
V
i
r
t
ua
l
s
c
r
e
e
ni
ng
of
I
ndone
s
i
a
n
he
r
ba
l
c
om
pounds
a
s
C
O
V
I
D
-
19
s
uppor
t
i
ve
t
he
r
a
py:
m
a
c
hi
ne
l
e
a
r
ni
ng
a
nd
pha
r
m
a
c
ophor
e
m
ode
l
i
ng
a
ppr
oa
c
he
s
,”
B
M
C
C
om
pl
e
m
e
nt
ar
y
M
e
di
c
i
ne
and
T
he
r
api
e
s
,
vol
.
22,
no.
1,
D
e
c
.
2022,
doi
:
10.1186/
s
12906
-
022
-
03686
-
y.
[
12]
N
.
S
.
R
a
m
a
dha
nt
i
,
W
.
A
.
K
us
um
a
,
I
.
B
a
t
uba
r
a
,
a
nd
R
.
H
e
r
ya
nt
o,
“
R
a
ndom
f
or
e
s
t
t
o
pr
e
di
c
t
e
uc
a
l
ypt
us
a
s
a
pot
e
nt
i
a
l
h
e
r
b
i
n
pr
e
ve
nt
i
ng
C
O
V
I
D
19
,”
i
n
2021
I
E
E
E
C
onf
e
r
e
nc
e
on
C
om
put
at
i
onal
I
nt
e
l
l
i
ge
nc
e
i
n
B
i
oi
nf
or
m
at
i
c
s
and
C
om
put
at
i
onal
B
i
ol
ogy
(
C
I
B
C
B
)
,
I
E
E
E
, O
c
t
. 2021, pp. 01
–
05
, doi
:
10.1109/
C
I
B
C
B
49929.2021.9562940.
[
13]
A
.
A
.
M
a
l
i
k,
C
.
P
ha
nus
‐
um
por
n,
N
.
S
c
ha
dua
ngr
a
t
,
W
.
S
hoom
bua
t
ong,
C
.
I
s
a
r
a
nkur
a
‐
N
a
‐
A
yudhya
,
a
nd
C
.
N
a
nt
a
s
e
na
m
a
t
,
“
H
C
V
pr
e
d:
a
w
e
b
s
e
r
ve
r
f
or
pr
e
di
c
t
i
ng t
he
bi
oa
c
t
i
vi
t
y of
he
pa
t
i
t
i
s
C
vi
r
us
N
S
5B
i
nhi
bi
t
or
s
,”
J
our
nal
of
C
om
put
at
i
onal
C
he
m
i
s
t
r
y
,
vol
. 41, no. 20, pp. 1820
–
1834, J
ul
. 2020, doi
:
10.1002/
j
c
c
.26223.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
3003
-
3013
3012
[
14]
T
.
L
e
r
ks
ut
hi
r
a
t
,
S
.
C
hi
t
phuk,
W
.
S
t
i
t
c
ha
nt
r
a
kul
,
D
.
D
e
j
s
uphong,
A
.
A
.
M
a
l
i
k,
a
nd
C
.
N
a
nt
a
s
e
n
a
m
a
t
,
“
P
A
R
P
1pr
e
d:
a
w
e
b
s
e
r
ve
r
f
or
s
c
r
e
e
ni
ng
t
he
bi
oa
c
t
i
vi
t
y
of
i
nhi
bi
t
or
s
a
ga
i
ns
t
D
N
A
r
e
pa
i
r
e
nz
ym
e
P
A
R
P
-
1,”
E
X
C
L
I
J
our
nal
,
vol
.
22,
pp.
84
–
107,
2023,
doi
:
10.17179/
e
xc
l
i
2022
-
5602.
[
15]
C
.
B
e
nne
t
t
e
a
nd
A
.
V
i
c
ke
r
s
,
“
A
g
a
i
ns
t
qua
nt
i
l
e
s
:
c
a
t
e
gor
i
z
a
t
i
on
of
c
ont
i
nu
ous
va
r
i
a
bl
e
s
i
n
e
pi
de
m
i
ol
ogi
c
r
e
s
e
a
r
c
h,
a
nd
i
t
s
di
s
c
ont
e
nt
s
,”
B
M
C
M
e
di
c
al
R
e
s
e
ar
c
h M
e
t
hodol
ogy
, vol
. 12, no. 1, D
e
c
. 2012, d
oi
:
10.1186/
1471
-
2288
-
12
-
21.
[
16]
K
.
E
.
M
a
r
kon,
M
.
C
hm
i
e
l
e
w
s
ki
,
a
nd
C
.
J
.
M
i
l
l
e
r
,
“
T
he
r
e
l
i
a
bi
l
i
t
y
a
nd
va
l
i
di
t
y
of
d
i
s
c
r
e
t
e
a
nd
c
ont
i
nuous
m
e
a
s
ur
e
s
of
ps
yc
hopa
t
hol
ogy:
a
qua
nt
i
t
a
t
i
ve
r
e
vi
e
w
,”
P
s
y
c
hol
ogi
c
al
B
ul
l
e
t
i
n
, vol
. 137, no.
5, pp. 856
–
879, 2011, doi
:
10.1037/
a
0023678.
[
17]
Z
.
G
a
o,
R
.
X
i
a
,
a
nd
P
.
Z
ha
ng,
“
P
r
e
di
c
t
i
on
of
a
nt
i
-
pr
ol
i
f
e
r
a
t
i
on
e
f
f
e
c
t
of
[
1,2,3]
T
r
i
a
z
ol
o
[
4,5
-
d]
pyr
i
m
i
di
ne
de
r
i
va
t
i
ve
s
by
r
a
ndom
f
or
e
s
t
a
nd
m
i
x
-
ke
r
ne
l
f
unc
t
i
on
S
V
M
w
i
t
h
P
S
O
,”
C
he
m
i
c
al
and
P
har
m
ac
e
ut
i
c
al
B
ul
l
e
t
i
n
,
vol
.
70,
no.
10,
O
c
t
.
2022,
doi
:
10.1248/
c
pb.c
22
-
00376.
[
18]
S
. K
a
m
boj
, A
. R
a
j
put
, A
. R
a
s
t
ogi
, A
. T
ha
kur
, a
nd
M
. K
um
a
r
, “
T
a
r
ge
t
i
ng non
-
s
t
r
uc
t
ur
a
l
pr
ot
e
i
ns
of
he
pa
t
i
t
i
s
C
vi
r
us
f
or
pr
e
di
c
t
i
ng
r
e
pur
pos
e
d
dr
ugs
us
i
ng
Q
S
A
R
a
nd
m
a
c
hi
ne
l
e
a
r
ni
ng
a
ppr
oa
c
he
s
,”
C
om
put
at
i
o
nal
and
St
r
uc
t
ur
al
B
i
ot
e
c
hnol
ogy
J
ou
r
nal
,
vol
.
20,
pp. 3422
–
3438, 2022, doi
:
10.1016/
j
.c
s
bj
.2022.06.060.
[
19]
D
.
N
.
F
i
a
t
e
t
al
.
,
“
C
om
pa
r
a
t
i
ve
a
na
l
ys
i
s
of
H
e
pa
t
i
t
i
s
C
vi
r
us
ge
not
ype
1a
(
I
s
ol
a
t
e
1)
us
i
ng
m
ul
t
i
pl
e
r
e
gr
e
s
s
i
on
a
l
gor
i
t
hm
s
a
nd
f
i
nge
r
pr
i
nt
i
ng
t
e
c
hni
que
s
,”
J
our
nal
of
E
l
e
c
t
r
oni
c
s
,
E
l
e
c
t
r
o
m
e
di
c
al
E
ngi
ne
e
r
i
ng,
and
M
e
di
c
al
I
nf
or
m
at
i
c
s
,
vol
.
6,
no.
4,
pp. 478
–
488, S
e
p. 2024, doi
:
10.35882/
j
e
e
e
m
i
.v6i
4.506.
[
20]
T
.
R
.
N
ovi
a
ndy,
G
.
M
.
I
dr
oe
s
,
T
.
E
.
T
a
l
l
e
i
,
D
.
H
a
nda
ya
ni
,
a
nd
R
.
I
dr
oe
s
,
“
Q
S
A
R
m
ode
l
i
ng
f
or
pr
e
di
c
t
i
ng
be
t
a
-
s
e
c
r
e
t
a
s
e
1
i
nhi
bi
t
or
y
a
c
t
i
vi
t
y
i
n
a
l
z
he
i
m
e
r
’
s
di
s
e
a
s
e
w
i
t
h
s
uppor
t
ve
c
t
or
r
e
gr
e
s
s
i
on,”
M
al
ac
c
a
P
har
m
ac
e
ut
i
c
s
,
vol
.
2,
no.
2,
pp.
79
–
85,
S
e
p. 2024, doi
:
10.60084/
m
p.v2i
2.226.
[
21]
G
.
I
.
B
.
J
a
na
i
r
o,
D
.
E
.
C
.
Y
u,
a
nd
J
.
I
.
B
. J
a
na
i
r
o,
“
A
m
a
c
hi
ne
l
e
a
r
ni
ng
r
e
gr
e
s
s
i
on
m
ode
l
f
or
t
he
s
c
r
e
e
ni
ng
a
nd
de
s
i
gn
of
pot
e
nt
i
a
l
S
A
R
S
-
C
oV
-
2
pr
ot
e
a
s
e
i
nhi
bi
t
or
s
,”
N
e
t
w
or
k
M
ode
l
i
ng
A
nal
y
s
i
s
i
n
H
e
al
t
h
I
nf
or
m
at
i
c
s
and
B
i
oi
nf
or
m
at
i
c
s
,
vol
.
10,
no.
1,
D
e
c
. 2021, doi
:
10.1007/
s
13721
-
021
-
00326
-
2.
[
22]
D
.
F
.
S
e
ngke
y
a
nd
A
.
M
a
s
e
ngi
,
“
R
e
gr
e
s
s
i
on
a
l
gor
i
t
hm
s
i
n
pr
e
di
c
t
i
ng
t
he
S
A
R
S
-
C
oV
-
2
r
e
pl
i
c
a
s
e
pol
ypr
ot
e
i
n
1a
b
i
nhi
bi
t
or
:
a
c
om
pa
r
a
t
i
ve
s
t
udy,”
J
our
nal
of
E
l
e
c
t
r
oni
c
s
,
E
l
e
c
t
r
om
e
di
c
al
E
ngi
ne
e
r
i
ng,
a
nd
M
e
di
c
al
I
nf
or
m
at
i
c
s
,
vol
.
6,
no.
1,
pp.
1
–
10,
D
e
c
. 2023, doi
:
10.35882/
j
e
e
e
m
i
.v6i
1.338.
[
23]
Y
.
C
á
r
de
na
s
‐
C
on
e
j
o,
A
.
L
i
ñ
a
n‐
R
i
c
o,
D
.
A
.
G
a
r
c
í
a
‐
R
odr
í
gue
z
,
S
.
C
e
nt
e
no‐
L
e
i
j
a
,
a
nd
H
.
S
e
r
r
a
no‐
P
os
a
da
,
“
A
n
e
xc
l
us
i
ve
42
a
m
i
no
a
c
i
d
s
i
gna
t
ur
e
i
n
pp1a
b
pr
ot
e
i
n
pr
ovi
de
s
i
ns
i
ght
s
i
nt
o
t
he
e
vol
ut
i
ve
hi
s
t
or
y
of
t
he
2019
nove
l
hum
a
n‐
pa
t
hoge
ni
c
c
or
ona
vi
r
u
s
(
S
A
R
S
‐
C
oV
‐
2)
,”
J
our
nal
of
M
e
di
c
al
V
i
r
ol
ogy
, vol
. 92, no. 6, pp. 688
–
692, J
un
. 2020, doi
:
10.1002/
j
m
v.25758.
[
24]
R
.
Y
a
da
v
e
t
al
.
,
“
R
ol
e
of
s
t
r
uc
t
ur
a
l
a
nd
non
-
s
t
r
uc
t
ur
a
l
pr
ot
e
i
ns
a
nd
t
h
e
r
a
pe
ut
i
c
t
a
r
ge
t
s
of
S
A
R
S
-
C
oV
-
2
f
or
C
O
V
I
D
-
19,”
C
e
l
l
s
,
vol
. 10, no. 4, A
pr
. 2021, doi
:
10.3390/
c
e
l
l
s
10040821.
[
25]
T
.
B
a
dr
i
ya
h,
D
.
B
.
S
a
nt
o
s
o,
I
.
S
ya
r
i
f
,
a
nd
D
.
R
.
S
ya
r
i
f
,
“
I
m
pr
ovi
ng
s
t
r
oke
di
a
gnos
i
s
a
c
c
ur
a
c
y
u
s
i
ng
hype
r
pa
r
a
m
e
t
e
r
opt
i
m
i
z
e
d
de
e
p
l
e
a
r
ni
ng,”
I
nt
e
r
nat
i
onal
J
our
nal
of
A
dv
anc
e
s
i
n
I
nt
e
l
l
i
ge
nt
I
nf
or
m
at
i
c
s
,
vol
.
5,
no.
3,
N
ov.
2019,
doi
:
10.26555/
i
j
a
i
n.v5i
3.427.
[
26]
H
.
J
.
P
.
W
e
e
r
t
s
,
A
.
C
.
M
ue
l
l
e
r
,
a
nd
J
.
V
a
ns
c
hor
e
n,
“
I
m
por
t
a
nc
e
of
t
uni
ng
hype
r
pa
r
a
m
e
t
e
r
s
of
m
a
c
hi
ne
l
e
a
r
ni
ng
a
l
gor
i
t
hm
s
,”
ar
X
i
v
-
C
om
put
e
r
Sc
i
e
nc
e
,
pp. 1
-
17, J
ul
.
2020
.
[
27]
B
.
B
i
s
c
hl
e
t
al
.
,
“
H
yp
e
r
pa
r
a
m
e
t
e
r
opt
i
m
i
z
a
t
i
on:
f
ounda
t
i
ons
,
a
l
gor
i
t
hm
s
,
be
s
t
pr
a
c
t
i
c
e
s
,
a
nd
op
e
n
c
ha
l
l
e
nge
s
,”
W
i
l
e
y
I
nt
e
r
di
s
c
i
pl
i
nar
y
R
e
v
i
e
w
s
:
D
at
a M
i
ni
ng and K
now
l
e
dge
D
i
s
c
ov
e
r
y
, vol
. 13, no.
2, 2023, doi
:
10.1002/
w
i
dm
.1484.
[
28]
A
.
B
oul
e
s
t
e
i
x,
B
.
B
i
s
c
hl
,
a
nd
P
.
P
r
obs
t
,
“
T
una
bi
l
i
t
y:
i
m
por
t
a
nc
e
of
hype
r
pa
r
a
m
e
t
e
r
s
of
m
a
c
hi
ne
l
e
a
r
ni
ng
a
l
gor
i
t
hm
s
,”
J
our
nal
of
M
ac
hi
ne
L
e
ar
ni
ng R
e
s
e
ar
c
h
, vol
. 20, no. 53, 2019.
[
29]
F
.
P
e
dr
e
gos
a
e
t
al
.
,
“
S
c
i
ki
t
-
l
e
a
r
n:
m
a
c
hi
ne
l
e
a
r
ni
ng
i
n
P
yt
hon,”
J
ou
r
nal
of
M
ac
hi
ne
L
e
ar
ni
ng
R
e
s
e
ar
c
h
,
vol
.
127,
no.
9,
pp. 2825
–
2830, 2019.
[
30]
M
.
D
a
vi
e
s
e
t
al
.
,
“
C
hE
M
B
L
w
e
b
s
e
r
vi
c
e
s
:
s
t
r
e
a
m
l
i
ni
ng
a
c
c
e
s
s
t
o
dr
ug
di
s
c
ove
r
y
da
t
a
a
nd
ut
i
l
i
t
i
e
s
,”
N
uc
l
e
i
c
A
c
i
ds
R
e
s
e
ar
c
h
,
vol
. 43, no. W
1, J
ul
. 2015, doi
:
10.1093/
na
r
/
gkv352.
[
31]
C
.
W
.
Y
a
p,
“
P
a
D
E
L
‐
de
s
c
r
i
pt
or
:
a
n
ope
n
s
our
c
e
s
of
t
w
a
r
e
t
o
c
a
l
c
ul
a
t
e
m
o
l
e
c
ul
a
r
de
s
c
r
i
pt
or
s
a
nd
f
i
nge
r
pr
i
nt
s
,”
J
our
nal
of
C
om
put
at
i
onal
C
he
m
i
s
t
r
y
, vol
. 32, no. 7, pp. 1466
–
1474, M
a
y 2011, doi
:
10.10
02/
j
c
c
.21707.
[
32]
R
C
or
e
T
e
a
m
,
R
:
a
l
anguage
and
e
nv
i
r
onm
e
nt
f
or
s
t
at
i
s
t
i
c
al
c
om
put
i
ng
,
R
F
ou
nda
t
i
on
f
or
S
t
a
t
i
s
t
i
c
a
l
C
om
put
i
ng
,
V
i
e
nna
,
A
us
t
r
i
a
,
2024. [
O
nl
i
ne
]
. A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
c
r
a
n.r
-
pr
oj
e
c
t
.or
g/
doc
/
m
a
nua
l
s
/
r
-
r
e
l
e
a
s
e
/
f
ul
l
r
e
f
m
a
n.pdf
[
33]
N
. N
. I
va
nov, D
. A
. S
hul
ga
, a
nd V
. A
. P
a
l
yul
i
n, “
D
e
c
om
po
s
i
t
i
on of
s
m
a
l
l
m
ol
e
c
ul
e
s
f
or
f
r
a
gm
e
nt
-
ba
s
e
d dr
ug de
s
i
gn,”
B
i
ophy
s
i
c
a
,
vol
. 3, no. 2, pp. 362
–
372, 2023, doi
:
10.3390/
bi
ophys
i
c
a
3020024.
[
34]
G
.
W
.
B
e
m
i
s
a
nd
M
.
A
.
M
ur
c
ko,
“
T
he
pr
ope
r
t
i
e
s
of
know
n
d
r
ugs
.
1.
m
ol
e
c
u
l
a
r
f
r
a
m
e
w
or
ks
,”
J
our
nal
of
M
e
di
c
i
nal
C
he
m
i
s
t
r
y
,
vol
. 39, no. 15, pp. 2887
–
2893, J
a
n. 1996, doi
:
10.1021/
j
m
9602928.
[
35]
A
.
K
um
a
r
,
S
.
L
oha
r
c
h,
S
.
K
um
a
r
,
R
.
P
.
R
i
ng
e
,
a
nd
R
.
P
a
r
ke
s
h,
“
E
xpl
oi
t
i
ng
c
h
e
m
i
nf
or
m
a
t
i
c
a
nd
m
a
c
hi
ne
l
e
a
r
ni
ng
t
o
na
vi
ga
t
e
t
he
a
va
i
l
a
bl
e
c
he
m
i
c
a
l
s
pa
c
e
of
pot
e
nt
i
a
l
s
m
a
l
l
m
ol
e
c
ul
e
i
nhi
bi
t
or
s
of
S
A
R
S
-
C
o
V
-
2,”
C
om
put
at
i
onal
and
St
r
uc
t
ur
al
B
i
ot
e
c
hnol
ogy
J
our
nal
, vol
. 19, pp. 424
–
438, 2021, doi
:
10.1016/
j
.c
s
bj
.2020.12.028.
[
36]
T
.
Y
u
e
t
a
l
.
,
“
E
xpl
o
r
i
ng
t
he
c
he
m
i
c
a
l
s
pa
c
e
of
C
Y
P
17A
1
i
nhi
bi
t
or
s
us
i
ng
c
h
e
m
i
nf
or
m
a
t
i
c
s
a
nd
m
a
c
hi
ne
l
e
a
r
ni
ng,”
M
o
l
e
c
ul
e
s
,
vol
. 28, no. 4, F
e
b. 2023, doi
:
10.3390/
m
ol
e
c
ul
e
s
28041679.
[
37]
R
.
G
uha
,
“
C
h
e
m
i
c
a
l
i
nf
or
m
a
t
i
c
s
f
unc
t
i
ona
l
i
t
y
i
n
R
,”
J
ou
r
nal
of
St
at
i
s
t
i
c
al
Sof
t
w
ar
e
,
vol
.
18,
no.
5,
2007,
doi
:
10.18637/
j
s
s
.v018.i
05.
[
38]
K
. S
. N
ugr
oho, A
. Y
. S
ukm
a
de
w
a
, A
.
V
i
di
a
nt
o, a
nd W
. F
.
M
a
hm
udy, “
E
f
f
e
c
t
i
ve
pr
e
di
c
t
i
ve
m
ode
l
l
i
ng f
or
c
or
ona
r
y a
r
t
e
r
y di
s
e
a
s
e
s
us
i
ng
s
uppor
t
ve
c
t
or
m
a
c
hi
ne
,”
I
A
E
S
I
nt
e
r
nat
i
onal
J
our
nal
of
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
(
I
J
-
A
I
)
,
vol
.
11,
no.
1,
M
a
r
.
2022,
doi
:
10.11591/
i
j
a
i
.v11.i
1.pp345
-
355.
[
39]
P
.
S
c
hr
a
t
z
,
J
.
M
ue
nc
how
,
E
.
I
t
u
r
r
i
t
xa
,
J
.
R
i
c
ht
e
r
,
a
nd
A
.
B
r
e
nni
ng,
“
H
ype
r
pa
r
a
m
e
t
e
r
t
uni
ng
a
nd
pe
r
f
or
m
a
nc
e
a
s
s
e
s
s
m
e
nt
of
s
t
a
t
i
s
t
i
c
a
l
a
nd
m
a
c
hi
ne
-
l
e
a
r
ni
ng
a
l
gor
i
t
hm
s
us
i
ng
s
pa
t
i
a
l
da
t
a
,”
E
c
ol
ogi
c
al
M
ode
l
l
i
ng
,
vol
.
406,
pp.
109
–
120,
A
ug.
2019,
doi
:
10.1016/
j
.e
c
ol
m
ode
l
.2019.06.002.
[
40]
M
.
S
i
ppe
r
,
“
H
i
gh
pe
r
pa
r
a
m
e
t
e
r
:
a
l
a
r
ge
-
s
c
a
l
e
s
t
udy
of
hype
r
pa
r
a
m
e
t
e
r
t
uni
ng
f
or
m
a
c
hi
ne
l
e
a
r
ni
ng
a
l
gor
i
t
hm
s
,”
A
l
gor
i
t
hm
s
,
vol
. 15, no. 9, S
e
p. 2022, doi
:
10.3390/
a
15090315.
[
41]
A
.
R
á
c
z
,
D
.
B
a
j
u
s
z
,
a
nd
K
.
H
é
be
r
ge
r
,
“
E
f
f
e
c
t
of
da
t
a
s
e
t
s
i
z
e
a
nd
t
r
a
i
n/
t
e
s
t
s
pl
i
t
r
a
t
i
os
i
n
Q
S
A
R
/
Q
S
P
R
m
ul
t
i
c
l
a
s
s
c
l
a
s
s
i
f
i
c
a
t
i
on,
”
M
ol
e
c
ul
e
s
, vol
. 26, no. 4, F
e
b. 2021, doi
:
10.3390/
m
ol
e
c
ul
e
s
26041111.
[
42]
D
.
M
oha
pa
t
r
a
,
S
.
K
.
B
hoi
,
C
.
M
a
l
l
i
c
k,
K
.
K
.
J
e
na
,
a
nd
S
.
M
i
s
hr
a
,
“
D
i
s
t
r
i
but
i
on
pr
e
s
e
r
vi
ng
t
r
a
i
n
-
t
e
s
t
s
pl
i
t
di
r
e
c
t
e
d
e
ns
e
m
bl
e
c
l
a
s
s
i
f
i
e
r
f
or
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
t
i
on,”
I
nt
e
r
nat
i
onal
J
our
nal
of
I
nf
or
m
at
i
on
T
e
c
hnol
ogy
,
vol
.
14,
no.
4,
pp.
1763
–
1769,
J
un. 2022, doi
:
10.1007/
s
41870
-
022
-
00868
-
2.
[
43]
J
. J
. S
a
l
a
z
a
r
, L
. G
a
r
l
a
nd, J
. O
c
hoa
, a
nd
M
. J
. P
yr
c
z
,
“
F
a
i
r
t
r
a
i
n
-
t
e
s
t
s
pl
i
t
i
n m
a
c
hi
ne
l
e
a
r
ni
ng:
m
i
t
i
ga
t
i
ng s
pa
t
i
a
l
a
ut
oc
or
r
e
l
a
t
i
on f
or
i
m
pr
ove
d
pr
e
di
c
t
i
on
a
c
c
ur
a
c
y,”
J
our
nal
of
P
e
t
r
ol
e
um
Sc
i
e
nc
e
and
E
ngi
ne
e
r
i
ng
,
vol
.
209,
F
e
b.
2022,
doi
:
10.1016/
j
.pe
t
r
ol
.2021.109885.
Evaluation Warning : The document was created with Spire.PDF for Python.