bioRxiv (Cold Spring Harbor Laboratory),
Journal Year:
2024,
Volume and Issue:
unknown
Published: Dec. 17, 2024
Abstract
Visual
interneurons
come
in
many
different
flavors,
representing
luminance
changes
at
one
location
as
ON
or
OFF
signals
with
dynamics,
ranging
from
purely
sustained
to
sharply
transient
responses.
While
the
functional
relevance
of
this
representation
for
subsequent
computations
like
direction-selective
motion
detection
is
well
understood,
mechanisms
by
which
these
differences
dynamics
arise
are
unclear.
Here,
I
study
question
fly
optic
lobe.
Taking
advantage
known
connectome
simulate
a
network
five
adjacent
optical
columns
each
comprising
65
cell
types.
Each
neuron
modeled
an
electrically
compact
single
compartment,
conductance-based
element
that
receives
input
other
neurons
within
its
column
and
neighboring
according
intra-
inter-columnar
connectivity
matrix.
The
sign
determined
transmitter
type
presynaptic
receptor
on
postsynaptic
side.
In
addition,
some
given
voltage-dependent
conductances
transcriptome.
As
free
parameters,
has
output
gain,
applied
all
synapses,
respectively.
parameters
adjusted
such
spatio-temporal
receptive
field
properties
13
out
simulated
match
experimentally
ones
closely
possible.
Despite
fact
have
identical
leak
conductance
membrane
capacitance,
procedure
leads
surprisingly
good
fit
data,
where
specific
respond
transiently
while
others
way
changes.
This
critically
depends
presence
H-current
first-order
interneurons,
i.e.,
lamina
cells
L1
L2:
turning
off
eliminates
response
nature
leaving
only
responses
examined
interneurons.
conclude
diverse
dynamic
behavior
columnar
lobe
starts
created
their
intrinsic
properties.
predict
eliminating
hyperpolarization-activated
current
RNAi
should
strongly
affect
medulla
and,
consequently,
also
higher-order
functions
depending
them
direction-selectivity
T4
T5
neurons.
Author
summary
transient.
computations,
detection,
mechanism
using
simulating
biophysically
plausible
way.
Adjusting
gain
subset
(those
experimental
data
exist)
matches
kinetics
real
counterparts,
identify
particular
voltage-gated
ion
channel
present
being
critical
study,
therefore,
predicts
circuit
downstream
elements
destruct
computation
direction
selectivity.
Nature,
Journal Year:
2025,
Volume and Issue:
unknown
Published: March 26, 2025
Abstract
Vision
provides
animals
with
detailed
information
about
their
surroundings
and
conveys
diverse
features
such
as
colour,
form
movement
across
the
visual
scene.
Computing
these
parallel
spatial
requires
a
large
network
of
neurons.
Consequently,
from
flies
to
humans,
regions
in
brain
constitute
half
its
volume.
These
often
have
marked
structure–function
relationships,
neurons
organized
along
maps
shapes
that
directly
relate
roles
processing.
More
than
century
anatomical
studies
catalogued
detail
cell
types
fly
systems
1–3
,
behavioural
physiological
experiments
examined
capabilities
flies.
To
unravel
diversity
complex
system,
careful
mapping
neural
architecture
matched
tools
for
targeted
exploration
this
circuitry
is
essential.
Here
we
present
connectome
right
optic
lobe
male
Drosophila
melanogaster
acquired
using
focused
ion
beam
milling
scanning
electron
microscopy.
We
established
comprehensive
inventory
developed
computational
framework
quantify
anatomy.
Together,
data
establish
basis
interpreting
how
vision.
By
integrating
analysis
connectivity
information,
neurotransmitter
identity
expert
curation,
classified
approximately
53,000
into
732
types.
are
systematically
described
newly
named.
Finally,
share
an
extensive
collection
split-GAL4
lines
our
neuron-type
catalogue.
Overall,
set
unlocks
new
possibilities
systematic
investigations
vision
foundation
deeper
understanding
sensory
bioRxiv (Cold Spring Harbor Laboratory),
Journal Year:
2024,
Volume and Issue:
unknown
Published: March 14, 2024
Abstract
The
body
of
an
animal
influences
how
the
nervous
system
produces
behavior.
Therefore,
detailed
modeling
neural
control
sensorimotor
behavior
requires
a
model
body.
Here
we
contribute
anatomically-detailed
biomechanical
whole-body
fruit
fly
Drosophila
melanogaster
in
MuJoCo
physics
engine.
Our
is
general-purpose,
enabling
simulation
diverse
behaviors,
both
on
land
and
air.
We
demonstrate
generality
our
by
simulating
realistic
locomotion,
flight
walking.
To
support
these
have
extended
with
phenomenological
models
fluid
forces
adhesion
forces.
Through
data-driven
end-to-end
reinforcement
learning,
that
advances
enable
training
network
controllers
capable
locomotion
along
complex
trajectories
based
high-level
steering
signals.
use
visual
sensors
re-use
pre-trained
general-purpose
controller
to
perform
visually
guided
tasks.
project
open-source
platform
for
embodied
context.
bioRxiv (Cold Spring Harbor Laboratory),
Journal Year:
2023,
Volume and Issue:
unknown
Published: Oct. 13, 2023
A
catalog
of
neuronal
cell
types
has
often
been
called
a
"parts
list"
the
brain,
and
regarded
as
prerequisite
for
understanding
brain
function.
In
optic
lobe
Nature Neuroscience,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Nov. 22, 2024
A
long-standing
goal
in
neuroscience
is
to
understand
how
a
circuit's
form
influences
its
function.
Here,
we
reconstruct
and
analyze
synaptic
wiring
diagram
of
the
larval
zebrafish
brainstem
predict
key
functional
properties
validate
them
through
comparison
with
physiological
data.
We
identify
modules
strongly
connected
neurons
that
turn
out
be
specialized
for
different
behavioral
functions,
control
eye
body
movements.
The
movement
module
further
organized
into
two
three-block
cycles
support
positive
feedback
long
hypothesized
underlie
low-dimensional
attractor
dynamics
oculomotor
control.
construct
neural
network
model
based
directly
on
reconstructed
makes
predictions
cellular-resolution
coding
position
dynamics.
These
are
verified
statistically
calcium
imaging-based
activity
recordings.
This
work
demonstrates
connectome-based
brain
modeling
can
reveal
previously
unknown
anatomical
structure
circuit
provide
insights
linking
authors
determine
vertebrate
behaviorally
associated
modules.
this
connectome
predicts
imaging
