<s>
Patch-sequencing	B-Algorithm
(	O
patch-seq	O
)	O
is	O
a	O
method	O
designed	O
for	O
tackling	O
specific	O
problems	O
involved	O
in	O
characterizing	O
neurons	O
.	O
</s>
<s>
By	O
combining	O
multiple	O
data	O
modalities	O
such	O
as	O
electrophysiology	B-Application
,	O
sequencing	O
and	O
microscopy	O
,	O
Patch-seq	O
allows	O
for	O
neurons	O
to	O
be	O
characterized	O
in	O
multiple	O
ways	O
simultaneously	O
.	O
</s>
<s>
Patch-seq	O
is	O
a	O
specialized	O
form	O
of	O
the	O
patch	O
clamp	O
recording	O
technique	O
,	O
the	O
gold	O
standard	O
of	O
single	O
cell	O
electrophysiological	B-Application
studies	O
for	O
its	O
millisecond	O
resolution	O
of	O
cellular	O
electrophysiology	B-Application
,	O
its	O
ability	O
to	O
detect	O
currents	O
of	O
specific	O
ions	O
,	O
and	O
perhaps	O
most	O
importantly	O
for	O
its	O
ability	O
to	O
form	O
a	O
voltage	O
clamp	O
on	O
the	O
cell	O
membrane	O
.	O
</s>
<s>
This	O
clamp	O
is	O
critical	O
for	O
testing	O
voltage	O
dependent	O
opening	O
dynamics	O
of	O
ion	O
channels	O
which	O
determine	O
a	O
membrane	O
's	O
resting	O
potentials	O
and	O
action	B-Algorithm
potentials	I-Algorithm
.	O
</s>
<s>
Neuroscience	O
often	O
requires	O
concordant	O
observation	O
of	O
electrophysiological	B-Application
dynamics	O
and	O
behavior	O
to	O
test	O
the	O
relationship	O
between	O
psychological	O
theories	O
and	O
their	O
underlying	O
biology	O
.	O
</s>
<s>
While	O
a	O
neurons	O
electrical	O
properties	O
are	O
an	O
important	O
when	O
considering	O
defining	O
a	O
cell	O
type	O
its	O
morphology	O
,	O
types	O
of	O
neurotransmitters	O
released	O
,	O
neurotransmitter	O
receptors	O
expressed	O
at	O
synapses	B-Application
,	O
as	O
well	O
as	O
the	O
neuron	O
's	O
location	O
in	O
the	O
nervous	O
system	O
and	O
its	O
local	O
circuit	O
are	O
equally	O
important	O
.	O
</s>
<s>
Neurons	O
come	O
in	O
a	O
huge	O
diversity	O
of	O
shapes	O
with	O
many	O
differences	O
in	O
cell	O
bodies	O
(	O
soma	O
)	O
,	O
dendrites	O
,	O
and	O
axons	B-Algorithm
.	O
</s>
<s>
Likewise	O
the	O
targets	O
of	O
a	O
neurons	O
axon	B-Algorithm
determine	O
its	O
outputs	O
.	O
</s>
<s>
The	O
types	O
of	O
synapses	B-Application
formed	O
between	O
neurons	O
axons	B-Algorithm
and	O
dendrites	O
are	O
equally	O
important	O
as	O
well	O
.	O
</s>
<s>
Dendritic	O
geometry	O
can	O
influence	O
the	O
electrical	O
behavior	O
of	O
a	O
neurons	O
as	O
well	O
,	O
having	O
a	O
massive	O
influence	O
on	O
how	O
dendrites	O
process	O
input	O
in	O
the	O
form	O
of	O
postsynaptic	B-Algorithm
potentials	I-Algorithm
.	O
</s>
<s>
Linking	O
the	O
sequencing	O
information	O
to	O
a	O
neuronal	O
subtype	O
,	O
defined	O
previously	O
by	O
electrophysiological	B-Application
and	O
morphological	O
characteristics	O
is	O
a	O
slow	O
and	O
complicated	O
process	O
.	O
</s>
<s>
The	O
simultaneous	O
capture	O
and	O
integration	O
of	O
multiple	O
data	O
types	O
by	O
patch-seq	O
makes	O
it	O
ideal	O
for	O
neuronal	O
classification	O
,	O
uncovering	O
new	O
correlations	O
between	O
gene	O
expression	O
,	O
electrophysiological	B-Application
and	O
morphological	O
properties	O
and	O
neuronal	O
function	O
.	O
</s>
<s>
This	O
makes	O
patch-seq	O
a	O
truly	O
interdisciplinary	O
method	O
,	O
requiring	O
collaboration	O
between	O
specialists	O
in	O
electrophysiology	B-Application
,	O
sequencing	O
,	O
and	O
imaging	O
.	O
</s>
<s>
As	O
mention	O
previously	O
the	O
position	O
their	O
dendrites	O
and	O
axons	B-Algorithm
extend	O
into	O
as	O
well	O
as	O
the	O
neuron	O
's	O
position	O
with	O
a	O
brain	O
structure	O
is	O
incredibly	O
important	O
for	O
understanding	O
its	O
role	O
within	O
a	O
circuit	O
.	O
</s>
<s>
Features	O
of	O
the	O
cell	O
body	O
's	O
membrane	O
are	O
recorded	O
including	O
resting	O
membrane	O
potential	O
and	O
threshold	B-Algorithm
potential	I-Algorithm
.	O
</s>
<s>
Features	O
observed	O
from	O
generated	O
action	B-Algorithm
potentials	I-Algorithm
(	O
AP	O
)	O
such	O
as	O
AP	O
width	O
,	O
AP	O
amplitude	O
,	O
after-depolarization	O
and	O
after-hyperpolarization	O
amplitude	O
are	O
also	O
recorded	O
.	O
</s>
<s>
Cells	O
are	O
then	O
stimulated	O
by	O
ramp	O
and	O
square	O
currents	O
,	O
their	O
electrophysiological	B-Application
properties	O
are	O
recorded	O
and	O
measured	O
.	O
</s>
<s>
The	O
staining	O
by	O
biotin	O
is	O
performed	O
during	O
neuron	O
electrophysiological	B-Application
recording	O
.	O
</s>
<s>
Quality	O
of	O
the	O
outcome	O
depends	O
on	O
a	O
variety	O
of	O
factors	O
from	O
the	O
electrophysiological	B-Application
recording	O
time	O
to	O
the	O
nucleus	O
extraction	O
procedure	O
.	O
</s>
<s>
High	O
quality	O
if	O
somas	O
and	O
processes	O
are	O
fully	O
visible	O
and	O
proper	O
digital	O
reconstruction	O
is	O
possible	O
,	O
medium	O
quality	O
if	O
somas	O
and	O
some	O
processes	O
are	O
visible	O
but	O
are	O
nor	O
compatible	O
with	O
3D	O
reconstruction	O
,	O
insufficient	O
axon	B-Algorithm
where	O
dendrites	O
are	O
filled	O
with	O
biotin	O
but	O
axons	B-Algorithm
are	O
weakly	O
dyed	O
,	O
and	O
failed	O
fills	O
that	O
lack	O
soma	O
staining	O
likely	O
due	O
to	O
the	O
subside	O
of	O
the	O
structure	O
post	O
nuclear	O
extraction	O
.	O
</s>
<s>
Dimensionality	O
reduction	O
methods	O
such	O
as	O
T-distributed	B-Algorithm
stochastic	I-Algorithm
neighbor	I-Algorithm
embedding	I-Algorithm
or	O
uniform	O
manifold	O
approximation	O
and	O
projection	O
can	O
then	O
be	O
used	O
for	O
visualization	O
of	O
the	O
collected	O
data	O
's	O
position	O
on	O
a	O
reference	O
atlas	O
of	O
higher	O
quality	O
scRNA-seq	O
data	O
.	O
</s>
<s>
Machine	O
learning	O
can	O
be	O
applied	O
in	O
order	O
to	O
relate	O
the	O
gene	O
expression	O
data	O
to	O
the	O
morphological	O
and	O
electrophysiological	B-Application
data	O
.	O
</s>
<s>
Methods	O
for	O
doing	O
so	O
include	O
autoencoders	B-Algorithm
,	O
bottleneck	B-Architecture
networks	I-Architecture
,	O
or	O
other	O
rank	O
reduction	O
methods	O
.	O
</s>
<s>
Including	O
morphological	O
data	O
has	O
proven	O
to	O
be	O
challenging	O
as	O
it	O
is	O
a	O
computer	B-Application
vision	I-Application
task	O
,	O
a	O
notoriously	O
complicated	O
problem	O
in	O
machine	O
learning	O
.	O
</s>
<s>
It	O
is	O
difficult	O
to	O
represent	O
imaging	O
data	O
from	O
the	O
morphological	O
reconstructions	O
as	O
a	O
feature	B-Algorithm
vector	I-Algorithm
for	O
including	O
in	O
the	O
analysis	O
.	O
</s>
<s>
The	O
technique	O
is	O
particularly	O
well	O
suited	O
for	O
neuroscience	O
but	O
in	O
general	O
any	O
tissue	O
where	O
it	O
is	O
of	O
interest	O
to	O
simultaneous	O
know	O
electrophysiology	B-Application
,	O
morphology	O
,	O
and	O
transcriptomics	O
would	O
find	O
use	O
for	O
patch-seq	O
.	O
</s>
<s>
Patch-seq	O
can	O
provide	O
electrophysiological	B-Application
and	O
morphological	O
data	O
,	O
with	O
morphology	O
being	O
less	O
common	O
due	O
to	O
the	O
difficulty	O
in	O
achieving	O
high	O
quality	O
standards	O
.	O
</s>
<s>
Nevertheless	O
no	O
other	O
electrophysiological	B-Application
recording	O
's	O
technique	O
can	O
match	O
patch-seq	O
'	O
s	O
temporal	O
resolution	O
or	O
ability	O
to	O
characterize	O
specific	O
ionic	O
currents	O
nor	O
offer	O
the	O
capacity	O
to	O
produce	O
a	O
voltage	O
clamp	O
.	O
</s>
<s>
Slowly	O
automated	B-Algorithm
patch	I-Algorithm
clamp	I-Algorithm
is	O
beginning	O
to	O
become	O
the	O
norm	O
but	O
as	O
of	O
yet	O
has	O
not	O
reached	O
wide	O
adoption	O
.	O
</s>
