The
pharmaceutical
market
has
been
growing
rapidly,
but
concerns
about
energy
and
resource
sustainability
have
made
it
important
to
consider
the
economical
sustainable
aspects
of
discovering
functional
molecules
in
synthetic
chemistry.
One
main
challenges
traditional
chemical
synthesis
is
that
labor-intensive
generates
a
lot
waste
due
repetitive
reaction
manipulation.
To
address
this
issue,
paper
presents
robotic
end
effector
system
with
three
degrees
freedom
(DOF)
facilitate
automation
tasks
drug
discovery
workcell.
This
robotics
features
unique
remote
center
motion
(RCM)
spherical-linear
mechanism
novel
hollow
double
spring
vacuum
actuator
(HDSVA)
uses
soft
elastic
material
springs
for
actuation
structural
integrity.
covers
design,
kinematics,
system.
HDSVA
modeled
analytically
interaction
between
membrane
examined.
Through
kinematic
analysis,
simulation
results,
experimental
evaluations,
we
examine
capabilities
validate
feasibility
automated
stirring
tasks.
Journal of Computational Chemistry,
Год журнала:
2024,
Номер
45(13), С. 937 - 952
Опубликована: Янв. 4, 2024
Abstract
Design
of
new
drugs
is
a
challenging
process:
candidate
molecule
should
satisfy
multiple
conditions
to
act
properly
and
make
the
least
side‐effect—perfect
candidates
selectively
attach
influence
only
targets,
leaving
off‐targets
intact.
The
amount
experimental
data
about
various
properties
molecules
constantly
grows,
promoting
data‐driven
approaches.
However,
applicability
typical
predictive
machine
learning
techniques
can
be
substantially
limited
by
lack
particular
target.
For
example,
there
are
many
known
Thrombin
inhibitors
(acting
as
anticoagulants),
but
very
number
Protein
C
(coagulants).
In
this
study,
we
present
our
approach
suggest
inhibitor
building
an
effective
representation
chemical
space.
aim,
developed
deep
model—autoencoder,
trained
on
large
set
in
SMILES
format
map
Further,
applied
different
sampling
strategies
generate
novel
coagulant
candidates.
Symmetrically,
tested
anticoagulant
candidates,
where
were
able
predict
their
inhibition
towards
Thrombin.
We
also
compare
with
MegaMolBART—another
generative
model,
exploiting
similar
principles
navigation
BioMedInformatics,
Год журнала:
2022,
Номер
2(4), С. 603 - 624
Опубликована: Ноя. 12, 2022
The
adoption
of
“artificial
intelligence
(AI)
in
drug
discovery”,
where
AI
is
used
the
process
pharmaceutical
research
and
development,
progressing.
By
using
ability
to
large
amounts
data,
which
a
characteristic
AI,
achieving
advanced
data
analysis
inference,
there
are
benefits
such
as
shortening
development
time,
reducing
costs,
workload
researchers.
There
various
problems
but
following
two
issues
particularly
problematic:
(1)
yearly
increases
time
cost
drugs
(2)
difficulty
finding
highly
accurate
target
genes.
Therefore,
screening
simulation
expected.
Researchers
have
high
demands
for
collection
utilization
infrastructure
analysis.
In
field
discovery,
example,
interest
use
with
amount
chemical
or
biological
available.
application
discovery
becoming
more
active
due
improvement
computer
processing
power
spread
machine-learning
frameworks,
including
deep
learning.
To
evaluate
performance,
statistical
indices
been
introduced.
However,
factors
affected
performance
not
revealed
completely.
this
study,
we
summarized
reviewed
applications
learning
BigData.
Understanding
drug-response
differences
in
cancer
treatments
is
one
of
the
most
challenging
aspects
personalized
medicine.
Recently,
graph
neural
networks
(GNNs)
have
become
state-of-the-art
methods
many
representation
learning
scenarios
bioinformatics.
However,
building
an
optimal
handcrafted
GNN
model
for
a
particular
drug
sensitivity
dataset
requires
manual
design
and
fine-tuning
hyperparameters
model,
which
time-consuming
expert
knowledge.In
this
work,
we
propose
AutoCDRP,
novel
framework
automated
predictor
using
GNNs.
Our
approach
leverages
surrogate
modeling
to
efficiently
search
effective
architecture.
AutoCDRP
uses
predict
performance
architectures
sampled
from
space,
allowing
it
select
architecture
based
on
evaluation
performance.
Hence,
can
identify
by
exploring
all
space.
Through
comprehensive
experiments
two
benchmark
datasets,
demonstrate
that
generated
surpasses
designs.
Notably,
identified
consistently
outperforms
best
baseline
first
epoch,
providing
further
evidence
its
effectiveness.https://github.com/BeObm/AutoCDRP.
CHINESE JOURNAL OF ANALYTICAL CHEMISTRY (CHINESE VERSION),
Год журнала:
2023,
Номер
51(10), С. 100315 - 100315
Опубликована: Сен. 22, 2023
Breast
cancer
is
one
of
the
most
common
cancers
and
topmost
cause
mortality
among
women
in
both
developed
developing
countries.
Currently
available
potent
drugs
for
breast
exhibit
adverse
effects,
which
may
be
caused
as
a
result
why
cancer-specific
are
found
to
ineffective
patients.
In
this
study,
we
exploited
interaction
six
potential
drug
compounds
(Bazedoxifene,
Exemestane,
Fulvestrant,
Raloxifene,
Tryprostatin
A,
Vorinostat)
with
three
associated
proteins
such
poly
(ADP-ribose)
polymerase-1;
PARP1
(PDB
ID:
5HA9)
cyclin-dependent
kinase
2;
CDK2
6GUE),
phosphatidylinositol
3-kinases
alpha;
PI3Kα
7K6O)
using
molecular
docking
studies.
Docking
results
indicate
that
Raloxifene
was
shown
inhibitor
5HA9
protein
two
hydrogen
bond
interactions
possesses
best
binding
affinity
-12.3
kcal/mol.
The
compound
Fulvestrant
shows
has
-10.2
kcal/mol
exhibits
6GUE
protein.
indicated
-10.6
showed
7K6O
interactions.
Molecular
dynamics
simulations
5HA9-Raloxifene,
6GUE-Fulvestrant,
7K6O-Raloxifene
were
executed
100
ns
through
root
mean
square
deviation
(RMSD),
fluctuation
(RMSF),
number
bonds,
radius
gyration,
energy
computed.
obtained
can
useful
treatment
cancer.
