Machine learning for synthetic gene circuit engineering
Current Opinion in Biotechnology,
Год журнала:
2025,
Номер
92, С. 103263 - 103263
Опубликована: Янв. 27, 2025
Язык: Английский
Integrating Engineered Living Materials with 3D Bioprinting
Advanced Functional Materials,
Год журнала:
2025,
Номер
unknown
Опубликована: Апрель 18, 2025
Abstract
Engineered
living
materials
(ELMs)
are
an
emerging
class
of
biohybrid
with
genetically
programmable
functionalities.
Integrating
ELMs
3D
bioprinting
synergizes
their
biological
programmability
the
geometry‐driven
functionality
3D‐printed
constructs,
transforming
these
into
practical
products
and
engineering
solutions.
This
integration
also
introduces
a
new
paradigm
in
additive
manufacturing
that
harnesses
“livingness”
encapsulated
microorganisms
as
active
element
fabrication
process
to
create
adaptive
evolving
constructs.
Perspective
presents
recent
advances
discusses
current
developments
at
intersection
ELMs.
It
highlights
opportunities
interface
two
fields,
including
understanding
interactions
between
nonliving
components
for
bioink
design,
incorporating
synthetic
biology
workflows,
utilizing
microbial
growth
postprinting
process,
integrating
shape‐morphing
enable
4D
printing
Язык: Английский
The Progress of Research on the Excavation of Plant Biosynthetic Pathways of Natural Products
Hans Journal of Agricultural Sciences,
Год журнала:
2024,
Номер
14(08), С. 953 - 962
Опубликована: Янв. 1, 2024
Язык: Английский
Decoding pattern formation rules by integrating mechanistic modeling and deep learning
bioRxiv (Cold Spring Harbor Laboratory),
Год журнала:
2024,
Номер
unknown
Опубликована: Сен. 2, 2024
Abstract
Predictive
programming
of
self-organized
pattern
formation
using
living
cells
is
challenging
in
major
part
due
to
the
difficulty
navigating
through
high-dimensional
design
space
effectively.
The
emergence
and
characteristics
patterns
are
highly
sensitive
both
system
environmental
parameters.
Often,
optimal
conditions
able
generate
represent
a
small
fraction
possible
space.
Furthermore,
experimental
generation
quantification
typically
labor
intensive
low
throughput,
making
it
impractical
optimize
solely
based
on
trials
errors.
To
this
end,
simulations
well-formulated
mechanistic
model
can
facilitate
identification
for
formation.
However,
even
moderately
complex
make
these
computationally
prohibitive
when
applied
large
parameter
In
study,
we
demonstrate
how
integrating
modeling
with
machine
learning
significantly
accelerate
exploration
patterning
circuits
aid
deriving
human-interpretable
rules.
We
apply
strategy
program
ring
Pseudomonas
aeruginosa
synthetic
gene
circuit.
Our
approach
involved
training
neural
network
simulated
data
predict
10
million
times
faster
than
model.
This
was
then
used
across
vast
array
combinations,
far
exceeding
size
dataset
what
feasible
alone.
By
doing
so,
identified
many
combinations
desirable
patterns,
which
still
an
extremely
explored
parametric
next
validate
top
candidates
identify
coarse-grained
rules
patterning.
experimentally
demonstrated
control
guided
by
learned
work
highlights
effectiveness
rational
engineering
dynamics
cells.
Язык: Английский
Automated Design of Oligopools and Rapid Analysis of Massively Parallel Barcoded Measurements
ACS Synthetic Biology,
Год журнала:
2024,
Номер
unknown
Опубликована: Дек. 6, 2024
Oligopool
synthesis
and
next-generation
sequencing
enable
the
construction
characterization
of
large
libraries
designed
genetic
parts
systems.
As
library
sizes
grow,
it
becomes
computationally
challenging
to
optimally
design
numbers
primer
binding
sites,
barcode
sequences,
overlap
regions
obtain
efficient
assemblies
precise
measurements.
We
present
Calculator,
an
end-to-end
suite
algorithms
data
structures
that
rapidly
designs
many
thousands
oligonucleotides
within
oligopool
analyzes
billions
barcoded
reads.
introduce
several
novel
concepts
greatly
increase
analysis
throughput,
including
orthogonally
symmetric
design,
adaptive
decision
trees
for
a
Scry
classifier,
read
packing.
demonstrate
Calculator's
capabilities
across
computational
benchmarks
real-data
projects,
over
four
million
highly
unique
compact
barcodes
in
1.2
h,
universal
sites
one
200-mer
oligos
15
min,
about
500
deep
reads
per
hour,
all
on
8-core
desktop
computer.
Overall,
Calculator
accelerates
creative
use
massively
parallel
experiments
by
eliminating
complexity
their
analysis.
Язык: Английский
Combinatorial Nanoparticle-Bound ssDNA Oligonucleotide Library Synthesized by Split-and-Pool Synthesis
ACS Applied Bio Materials,
Год журнала:
2024,
Номер
8(1), С. 844 - 853
Опубликована: Дек. 30, 2024
Synthetic
ssDNA
oligonucleotides
hold
great
potential
for
various
applications,
including
DNA
aptamers,
digital
data
storage,
origami,
and
synthetic
genomes.
In
these
contexts,
precise
control
over
the
synthesis
of
strands
is
essential
generating
combinatorial
sequences
with
user-defined
parameters.
Desired
features
creating
include
easy
manipulation
strands,
effective
detection
unique
sequences,
a
straightforward
mechanism
strand
elongation
termination.
this
study,
we
present
split-and-pool
method
on
nanoparticles,
enabling
creation
scalable
libraries.
Our
approach
involves
coupling
to
ligating
double-digested
fragments
orientation-specific
synthesis,
attaching
final
single-digested
fragment
ensure
We
assess
quality
our
by
characterizing
both
nanoparticles
used
as
solid
supports,
confirming
that
produces
scalable,
nanoparticle-bound
libraries
controllable
lengths.
Язык: Английский