A guide for active learning in synergistic drug discovery
Shuhui Wang,
No information about this author
Alexandre Allauzen,
No information about this author
Philippe Nghe
No information about this author
et al.
Scientific Reports,
Journal Year:
2025,
Volume and Issue:
15(1)
Published: Jan. 28, 2025
Synergistic
drug
combination
screening
is
a
promising
strategy
in
discovery,
but
it
involves
navigating
costly
and
complex
search
space.
While
AI,
particularly
deep
learning,
has
advanced
synergy
predictions,
its
effectiveness
limited
by
the
low
occurrence
of
synergistic
pairs.
Active
which
integrates
experimental
testing
into
learning
process,
been
proposed
to
address
this
challenge.
In
work,
we
explore
key
components
active
provide
recommendations
for
implementation.
We
find
that
molecular
encoding
impact
on
performance,
while
cellular
environment
features
significantly
enhance
predictions.
Additionally,
can
discover
60%
pairs
with
only
exploring
10%
combinatorial
The
yield
ratio
observed
be
even
higher
smaller
batch
sizes,
where
dynamic
tuning
exploration-exploitation
further
performance.
code
found
at
https://github.com/LBiophyEvo/DrugSynergy.
Language: Английский
ActinoMation: a literate programming approach for medium-throughput robotic conjugation of Streptomyces spp.
Tenna Alexiadis Møller,
No information about this author
Thomas Booth,
No information about this author
Simon J. Shaw
No information about this author
et al.
Synthetic and Systems Biotechnology,
Journal Year:
2025,
Volume and Issue:
10(2), P. 667 - 676
Published: March 11, 2025
The
genus
Streptomyces
are
valuable
producers
of
antibiotics
and
other
pharmaceutically
important
bioactive
compounds.
Advances
in
molecular
engineering
tools,
such
as
CRISPR,
have
provided
some
access
to
the
metabolic
potential
Streptomyces,
but
efficient
genetic
strains
is
hindered
by
laborious
slow
manual
transformation
protocols.
In
this
paper,
we
present
a
semi-automated
medium-throughput
workflow
for
introduction
recombinant
DNA
into
spp.
using
affordable
open-sourced
Opentrons
(OT-2)
robotics
platform.
To
increase
accessibility
provide
an
open-source
protocol-creator,
ActinoMation.
ActinoMation
literate
programming
environment
Python
Jupyter
Notebook.
We
validated
method
transforming
coelicolor
(M1152
M1146),
S.
albidoflavus
(J1047),
venezuelae
(DSM40230)
with
plasmids
pSETGUS
pIJ12551.
demonstrate
conjugation
efficiencies
3.33∗10-3/0.33
%
M1152
pIJ12551;
2.96∗10-3/0.29
M1146
1.21∗10-5/0.0012
J1047
4.70∗10-4/0.047
pIJ12551,
4.97∗10-2/4.97
DSM40230
6.13∗10-2/6.13
pIJ12551
false
positive
rate
between
8.33
54.54
%.
Automation
facilitates
streamlined
on
larger
scale
without
any
evident
loss
efficiency.
Language: Английский
Debottlenecking cytochrome P450-dependent metabolic pathways for the biosynthesis of commercial natural products
Natural Product Reports,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Jan. 1, 2024
Covering:
2016
to
the
end
of
2024This
highlight
article
aims
provide
a
perspective
on
challenges
that
novel
biotechnological
processes
face
in
biomanufacturing
natural
products
(NPs)
whose
biosynthesis
pathways
rely
cytochrome
P450
monooxygenases.
This
enzyme
superfamily
is
one
most
versatile
plethora
NPs
finding
use
across
food,
nutrition,
medicine,
chemical
and
cosmetics
industries.
These
enzymes
often
exhibit
excellent
regio-
stereoselectivity,
but
they
can
suffer
from
low
activity
instability,
which
are
serious
issues
impairing
development
high
performing
bioprocesses.
We
start
with
brief
introduction
industrial
biotechnology
importance
looking
for
alternative
means
producing
independently
unsustainable
fossil
fuels
or
plant
extractions.
then
discuss
implemented
solutions
during
commercial
NP
focusing
P450-dependent
steps
primarily
yeast
cell
factories.
Our
main
focus
encountered
when
utilizing
pathways,
how
protein
engineering
be
used
debottlenecking
them.
Finally,
we
briefly
touch
upon
artificial
intelligence
machine
learning
guiding
efforts.
Language: Английский
ActinoMation: a literate programming approach for medium-throughput robotic conjugation of Streptomyces spp.
Tenna Alexiadis Møller,
No information about this author
Thomas Booth,
No information about this author
Simon J. Shaw
No information about this author
et al.
bioRxiv (Cold Spring Harbor Laboratory),
Journal Year:
2024,
Volume and Issue:
unknown
Published: Dec. 9, 2024
Abstract
The
genus
Streptomyces
are
valuable
producers
of
antibiotics
and
other
pharmaceutically
important
bioactive
compounds.
Advances
in
molecular
engineering
tools,
such
as
CRISPR,
has
provided
some
access
to
the
metabolic
potential
,
but
efficient
genetic
strains
is
hindered
by
laborious
slow
manual
transformation
protocols.
In
this
paper,
we
present
a
semi-automated
medium-throughput
workflow
for
introduction
recombinant
DNA
into
spp.
using
affordable
open-sourced
Opentrons
(OT-2)
robotics
platform.
To
increase
accessibility
provide
an
open-source
protocol-creator,
ActinoMation.
ActinoMation
literate
programming
environment
Python
Jupyter
Notebook.
We
validated
method
transforming
coelicolor
(M1152
M1146),
S.
albidoflavus
(J1047),
venezuelae
(DSM40230)
with
plasmids
pSETGUS
pIJ12551.
demonstrate
conjugation
efficiencies
3.33*10
−3
M1152
pIJ12551;
2.96*10
M1146
1.21*10
−5
J1047
4.70*10
−4
pIJ12551,
4.97*10
−2
DSM40230
6.13*10
pIJ12551
false
positive
rate
between
8.33%
54.54%.
Automation
improves
consistency
when
handling
large
sample
sizes
that
facilitates
easy
reproducibility
on
larger
scale.
Language: Английский