ACS Applied Materials & Interfaces,
Journal Year:
2024,
Volume and Issue:
16(43), P. 58754 - 58763
Published: Oct. 21, 2024
Metal–organic
frameworks
(MOFs)
are
versatile
nanoporous
materials
for
a
wide
variety
of
important
applications.
Recently,
handful
MOFs
have
been
explored
the
storage
toxic
fluorinated
gases
(Keasler
et
al.
Science,
2023,
381,
1455),
yet
potential
great
number
such
an
environmentally
sustainable
application
has
not
thoroughly
investigated.
In
this
work,
we
apply
active
learning
(AL)
to
accelerate
discovery
hypothetical
(hMOFs)
that
can
efficiently
store
specific
gas,
namely,
vinylidene
fluoride
(VDF).
First,
force
field
was
developed
VDF
and
utilized
predict
working
capacities
(ΔN)
in
initial
data
set
4502
from
computation-ready
experimental
MOF
(CoRE-MOF)
database
successfully
underwent
featurization
grand-canonical
Monte
Carlo
simulations.
Next,
diversified
by
Greedy
sampling
unexplored
sample
space
119,387
hMOFs
ab
initio
REPEAT
charge
(ARC-MOF)
database.
A
budget
10,000
samples
(i.e.,
<10%
total
ARC-MOFs)
selected
train
random
forest
model.
Then,
ΔN
unlabeled
ARC-MOFs
were
predicted
top-performing
ones
validated
Integrating
with
stability
requirement,
mechanically
stable
finally
identified,
along
high
ΔN.
Furthermore,
Pareto–Frontier
analysis,
revealed
long
linear
linkers
enhance
ΔN,
while
bulkier
multiphenyl
or
interpenetrated
improve
mechanical
strength.
From
discover
AL
also
demonstrate
importance
integrating
identify
promising
practical
application.
Journal of Cheminformatics,
Journal Year:
2024,
Volume and Issue:
16(1)
Published: Feb. 21, 2024
REINVENT
4
is
a
modern
open-source
generative
AI
framework
for
the
design
of
small
molecules.
The
software
utilizes
recurrent
neural
networks
and
transformer
architectures
to
drive
molecule
generation.
These
generators
are
seamlessly
embedded
within
general
machine
learning
optimization
algorithms,
transfer
learning,
reinforcement
curriculum
learning.
enables
facilitates
de
novo
design,
R-group
replacement,
library
linker
scaffold
hopping
optimization.
This
contribution
gives
an
overview
describes
its
design.
Algorithms
their
applications
discussed
in
detail.
command
line
tool
which
reads
user
configuration
either
TOML
or
JSON
format.
aim
this
release
provide
reference
implementations
some
most
common
algorithms
based
An
additional
goal
with
create
education
future
innovation
molecular
available
from
https://github.com/MolecularAI/REINVENT4
released
under
permissive
Apache
2.0
license.
Scientific
contribution.
provides
implementation
where
also
being
used
production
support
in-house
drug
discovery
projects.
publication
one
code
full
documentation
thereof
will
increase
transparency
foster
innovation,
collaboration
education.
Journal of Chemical Information and Modeling,
Journal Year:
2024,
Volume and Issue:
64(3), P. 653 - 665
Published: Jan. 30, 2024
The
incredible
capabilities
of
generative
artificial
intelligence
models
have
inevitably
led
to
their
application
in
the
domain
drug
discovery.
Within
this
domain,
vastness
chemical
space
motivates
development
more
efficient
methods
for
identifying
regions
with
molecules
that
exhibit
desired
characteristics.
In
work,
we
present
a
computationally
active
learning
methodology
and
demonstrate
its
applicability
targeted
molecular
generation.
When
applied
c-Abl
kinase,
protein
FDA-approved
small-molecule
inhibitors,
model
learns
generate
similar
inhibitors
without
prior
knowledge
existence
even
reproduces
two
them
exactly.
We
also
show
is
effective
any
commercially
available
HNH
CRISPR-associated
9
(Cas9)
enzyme.
To
facilitate
implementation
reproducibility,
made
all
our
software
through
open-source
ChemSpaceAL
Python
package.
Chemical Science,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 1, 2025
Existing
approaches
to
consider
the
synthesizability
of
generated
molecules.
This
work
demonstrates
use
an
explicit
retrosynthesis
model
directly
as
optimization
objective.
Journal of Medicinal Chemistry,
Journal Year:
2024,
Volume and Issue:
67(21), P. 18633 - 18636
Published: Oct. 24, 2024
2024
has
been
an
exciting
year
for
computational
sciences,
with
the
Nobel
Prize
in
Physics
awarded
"artificial
neural
network"
and
Chemistry
presented
"protein
structure
prediction
design".
Given
rapid
advancements
Computer-Aided
Drug
Design
(CADD)
Artificial
Intelligence
Discovery
(AIDD),
a
document
summarizing
their
current
standing
future
directions
would
be
timely
relevant
to
readership
of
