Syn-MolOpt: a synthesis planning-driven molecular optimization method using data-derived functional reaction templates
Journal of Cheminformatics,
Год журнала:
2025,
Номер
17(1)
Опубликована: Март 2, 2025
Molecular
optimization
is
a
crucial
step
in
drug
development,
involving
structural
modifications
to
improve
the
desired
properties
of
candidates.
Although
many
deep-learning-based
molecular
algorithms
have
been
proposed
and
may
perform
well
on
benchmarks,
they
usually
do
not
pay
sufficient
attention
synthesizability
molecules,
resulting
optimized
compounds
difficult
be
synthesized.
To
address
this
issue,
we
first
developed
general
pipeline
capable
constructing
functional
reaction
template
library
specific
any
property
where
predictive
model
can
built.
Based
these
templates,
introduced
Syn-MolOpt,
synthesis
planning-oriented
method.
During
optimization,
templates
steer
process
towards
by
effectively
transforming
relevant
fragments.
In
four
diverse
tasks,
including
two
toxicity-related
(GSK3β-Mutagenicity
GSK3β-hERG)
metabolism-related
(GSK3β-CYP3A4
GSK3β-CYP2C19)
multi-property
optimizations,
Syn-MolOpt
outperformed
three
benchmark
models
(Modof,
HierG2G,
SynNet),
highlighting
its
efficacy
adaptability.
Additionally,
visualization
synthetic
routes
for
molecules
confirms
effectiveness
optimization.
Notably,
Syn-MolOpt's
robust
performance
scenarios
with
limited
scoring
accuracy
demonstrates
potential
real-world
applications.
By
considering
both
synthesizability,
promises
valuable
tool
Scientific
contribution
takes
into
account
synthesis,
allowing
design
property-specific
libraries
optimized,
providing
reference
while
optimizing
targeted
properties.
universal
workflow
makes
it
suitable
various
types
tasks.
Язык: Английский
Local reaction condition optimization via machine learning
Journal of Molecular Modeling,
Год журнала:
2025,
Номер
31(5)
Опубликована: Апрель 23, 2025
Язык: Английский
Assessing Apparent Equilibrium Concentrations in Cementation of Trace Pd, Pt, Au, and Rh from Nitrate Solutions Using Mg, Al, Fe, and Zn
Metals,
Год журнала:
2024,
Номер
14(9), С. 990 - 990
Опубликована: Авг. 30, 2024
This
study
explores
the
impact
of
nitrate
ions
on
efficiency
cementing
noble
metals
from
diluted
waste
solutions
at
a
temperature
30
°C.
The
research
involved
measuring
effectiveness
different
(such
as
Zn,
Al,
Mg,
and
Fe)
in
presence
by
assessing
change
metal
ion
concentrations
before
after
cementation
process
using
spectrometric
analysis.
Initial
ware
Pt
=
5
ppm,
Au
7.5
Pd
Rh
1
ppm.
Kinetic
studies
revealed
that
24
h
is
adequate
to
achieve
apparent
equilibrium
with
pH
2
M
content.
identified
significant
recovery
losses
for
gold
platinum
solutions,
underlining
necessity
nitrate-free
recycling.
Zinc
magnesium
were
effective
Rh,
while
aluminum
was
efficient
reduction
each
condition.
Complete
removal
not
achieved
any
tested
metal,
indicating
need
alternative
methods.
Язык: Английский
Target-triggered catalytic hairpin assembly of miR-21/155/1 coupled with dsDNA-reporter amplified detection for prediction of clinically significant coronary artery disease
Sensors and Actuators B Chemical,
Год журнала:
2024,
Номер
unknown, С. 137053 - 137053
Опубликована: Ноя. 1, 2024
Язык: Английский
Recent Advancements in the Application of Artificial Intelligence in Drug Molecular Generation and Synthesis Planning
Pharmaceutical Fronts,
Год журнала:
2024,
Номер
unknown
Опубликована: Дек. 2, 2024
Abstract
The
design
and
synthesis
of
drug
molecules
is
a
pivotal
stage
in
development
that
traditionally
requires
significant
investment
time
finances.
However,
the
integration
artificial
intelligence
(AI)
accelerates
identification
potential
candidates,
optimizes
process,
contributes
to
more
informed
decision-making.
application
AI
molecular
generation
changing
way
researchers
explore
chemical
space
novel
compounds.
It
process
discovery
materials
science,
enabling
rapid
exploration
vast
landscapes
for
promising
candidates
further
experimental
validation.
predicting
reaction
products
planning
automation
synthetic
chemistry
tasks,
supports
chemists
making
decisions
during
discovery.
This
paper
reviewed
recent
advances
two
interrelated
areas:
routes.
will
provide
insights
into
innovative
ways
which
transforming
traditional
approaches
predict
its
future
progress
these
key
fields.
Язык: Английский