Nanoparticles
(NPs)
have
been
extensively
researched
for
targeted
diagnostic
imaging
and
drug
delivery,
yet
their
clinical
translation
remains
limited,
with
only
a
few
achieving
Food
Drug
Administration
approval.
This
limited
success
is
primarily
due
to
challenges
in
precise
organ-
or
tissue-specific
targeting,
which
arise
from
off-target
tissue
accumulation
suboptimal
clearance
profiles.
Herein
we
examine
the
critical
role
of
physicochemical
properties,
including
size,
surface
charge,
shape,
elasticity,
hardness,
density,
governing
biodistribution,
targetability,
NPs.
We
highlight
recent
advancements
engineering
NPs
showcasing
both
significant
progress
remaining
field
nanomedicine.
Additionally,
discuss
emerging
tools
technologies
that
are
being
developed
address
these
challenges.
Based
on
insights
materials
science,
biomedical
engineering,
computational
biology,
research,
propose
key
design
considerations
next-generation
nanomedicines
enhanced
organ
selectivity.
Nature Communications,
Год журнала:
2024,
Номер
15(1)
Опубликована: Июль 26, 2024
Abstract
Ionizable
lipid
nanoparticles
(LNPs)
are
seeing
widespread
use
in
mRNA
delivery,
notably
SARS-CoV-2
vaccines.
However,
the
expansion
of
therapies
beyond
COVID-19
is
impeded
by
absence
LNPs
tailored
for
diverse
cell
types.
In
this
study,
we
present
AI-Guided
Lipid
Engineering
(AGILE)
platform,
a
synergistic
combination
deep
learning
and
combinatorial
chemistry.
AGILE
streamlines
ionizable
development
with
efficient
library
design,
silico
screening
via
neural
networks,
adaptability
to
lines.
Using
AGILE,
rapidly
synthesize,
evaluate
lipids
selecting
from
vast
library.
Intriguingly,
reveals
cell-specific
preferences
lipids,
indicating
tailoring
optimal
delivery
varying
These
highlight
AGILE’s
potential
expediting
customized
LNPs,
addressing
complex
needs
clinical
practice,
thereby
broadening
scope
efficacy
therapies.
Frontiers in Cellular and Infection Microbiology,
Год журнала:
2025,
Номер
14
Опубликована: Янв. 20, 2025
Messenger
RNA
(mRNA)
vaccines
offer
an
adaptable
and
scalable
platform
for
cancer
immunotherapy,
requiring
optimal
design
to
elicit
a
robust
targeted
immune
response.
Recent
advancements
in
bioinformatics
artificial
intelligence
(AI)
have
significantly
enhanced
the
design,
prediction,
optimization
of
mRNA
vaccines.
This
paper
reviews
technologies
that
streamline
vaccine
development,
from
genomic
sequencing
lipid
nanoparticle
(LNP)
formulation.
We
discuss
how
accurate
predictions
neoantigen
structures
guide
sequences
effectively
target
cells.
Furthermore,
we
examine
AI-driven
approaches
optimize
mRNA-LNP
formulations,
enhancing
delivery
stability.
These
technological
innovations
not
only
improve
but
also
enhance
pharmacokinetics
pharmacodynamics,
offering
promising
avenues
personalized
immunotherapy.
Journal of Controlled Release,
Год журнала:
2024,
Номер
374, С. 219 - 229
Опубликована: Авг. 16, 2024
Nanoparticles
(NPs)
can
be
designed
for
targeted
delivery
in
cancer
nanomedicine,
but
the
challenge
is
a
low
efficiency
(DE)
to
tumor
site.
Understanding
impact
of
NPs'
physicochemical
properties
on
target
tissue
distribution
and
DE
help
improve
design
nanomedicines.
Multiple
machine
learning
artificial
intelligence
models,
including
linear
regression,
support
vector
machine,
random
forest,
gradient
boosting,
deep
neural
networks
(DNN),
were
trained
validated
predict
based
therapeutic
strategies
with
dataset
from
Nano-Tumor
Database.
Compared
other
DNN
model
had
superior
predictions
tumors
major
tissues.
The
determination
coefficients
(R
Advanced Materials,
Год журнала:
2025,
Номер
unknown
Опубликована: Фев. 2, 2025
Abstract
Despite
improvements
in
cancer
survival
rates,
metastatic
and
surgery‐resistant
cancers,
such
as
pancreatic
cancer,
remain
challenging,
with
poor
prognoses
limited
treatment
options.
Enhancing
drug
bioavailability
tumors,
while
minimizing
off‐target
effects,
is
crucial.
Metal–organic
frameworks
(MOFs)
have
emerged
promising
delivery
vehicles
owing
to
their
high
loading
capacity,
biocompatibility,
functional
tunability.
However,
the
vast
chemical
diversity
of
MOFs
complicates
rational
design
biocompatible
materials.
This
study
employed
machine
learning
molecular
simulations
identify
suitable
for
encapsulating
gemcitabine,
paclitaxel,
SN‐38,
identified
PCN‐222
an
optimal
candidate.
Following
loading,
MOF
formulations
are
improved
colloidal
stability
biocompatibility.
In
vitro
studies
on
cell
lines
shown
cellular
internalization,
delayed
release.
Long‐term
tests
demonstrated
a
consistent
performance
over
12
months.
vivo
tumor‐bearing
mice
revealed
that
paclitaxel‐loaded
PCN‐222,
particularly
hydrogel
local
administration,
significantly
reduced
spread
tumor
growth
compared
free
drug.
These
findings
underscore
potential
effective
system
hard‐to‐treat
cancers.
Nature Communications,
Год журнала:
2024,
Номер
15(1)
Опубликована: Ноя. 18, 2024
Abstract
The
rise
of
rational
strategies
in
nanomedicine
development,
such
as
high-throughput
methods
and
computer-aided
techniques,
has
led
to
a
shift
the
design
discovery
patterns
nanomedicines
from
trial-and-error
mode
mode.
This
transition
facilitates
enhancement
efficiency
preclinical
pipeline
nanomaterials,
particularly
improving
hit
rate
nanomaterials
optimization
promising
candidates.
Herein,
we
describe
directed
evolution
driven
by
data
accelerate
with
high
delivery
efficiency.
Computer-aided
are
introduced
detail
one
cutting-edge
directions
for
development
nanomedicines.
Ultimately,
look
forward
expanding
tools
using
multidisciplinary
approaches.
Rational
may
potentially
boost
next-generation
Nature Communications,
Год журнала:
2024,
Номер
15(1)
Опубликована: Дек. 30, 2024
Lipid
nanoparticles
(LNPs)
have
proven
effective
in
mRNA
delivery,
as
evidenced
by
COVID-19
vaccines.
Its
key
ingredient,
ionizable
lipids,
is
traditionally
optimized
inefficient
and
costly
experimental
screening.
This
study
leverages
artificial
intelligence
(AI)
virtual
screening
to
facilitate
the
rational
design
of
lipids
predicting
two
properties
LNPs,
apparent
pKa
delivery
efficiency.
Nearly
20
million
were
evaluated
through
iterations
AI-driven
generation
screening,
yielding
three
six
new
molecules,
respectively.
In
mouse
test
validation,
one
lipid
from
initial
iteration,
featuring
a
benzene
ring,
demonstrated
performance
comparable
control
DLin-MC3-DMA
(MC3).
Notably,
all
second
iteration
equaled
or
outperformed
MC3,
with
exhibiting
efficacy
akin
superior
SM-102.
Furthermore,
AI
model
interpretable
structure-activity
relationships.
