A point cloud segmentation framework for image-based spatial transcriptomics
Communications Biology,
Год журнала:
2024,
Номер
7(1)
Опубликована: Июль 6, 2024
Abstract
Recent
progress
in
image-based
spatial
RNA
profiling
enables
to
spatially
resolve
tens
hundreds
of
distinct
species
with
high
resolution.
It
presents
new
avenues
for
comprehending
tissue
organization.
In
this
context,
the
ability
assign
detected
transcripts
individual
cells
is
crucial
downstream
analyses,
such
as
in-situ
cell
type
calling.
Yet,
accurate
segmentation
can
be
challenging
data,
particular
absence
a
high-quality
membrane
marker.
To
address
issue,
we
introduce
ComSeg,
algorithm
that
operates
directly
on
single
positions
and
does
not
come
implicit
or
explicit
priors
shape.
ComSeg
applicable
complex
tissues
arbitrary
shapes.
Through
comprehensive
evaluations
simulated
experimental
datasets,
show
outperforms
existing
state-of-the-art
methods
single-cell
available
documented
open
source
pip
package
at
https://github.com/fish-quant/ComSeg
.
Язык: Английский
Trem2-expressing multinucleated giant macrophages are a biomarker of good prognosis in head and neck squamous cell carcinoma
Cancer Discovery,
Год журнала:
2024,
Номер
14(12), С. 2352 - 2366
Опубликована: Сен. 13, 2024
Abstract
Patients
with
head
and
neck
squamous
cell
carcinomas
(HNSCC)
often
have
poor
outcomes
due
to
suboptimal
risk
management
treatment
strategies;
yet
integrating
novel
prognostic
biomarkers
into
clinical
practice
is
challenging.
Here,
we
report
the
presence
of
multinucleated
giant
cells
(MGC)—a
type
macrophages—in
tumors
from
patients
HNSCC,
which
are
associated
a
favorable
prognosis
in
treatment-naive
preoperative
chemotherapy–treated
patients.
Importantly,
MGC
density
increased
following
therapy,
suggesting
role
these
antitumoral
response.
To
enable
translation
as
marker,
developed
deep-learning
model
automate
its
quantification
on
routinely
stained
pathological
whole
slide
images.
Finally,
used
spatial
transcriptomic
proteomic
approaches
describe
MGC-related
tumor
microenvironment
observed
an
increase
central
memory
CD4
T
cells.
We
defined
MGC-specific
signature
resembling
TREM2-expressing
mononuclear
tumor-associated
macrophages,
colocalized
keratin
niches.
Significance:
Novel
individual
needed
guide
therapeutic
decisions
for
cancer.
first
time,
granulomas
macrophages
keratin-rich
niches,
biomarker
slides.
Язык: Английский
Towards deciphering the bone marrow microenvironment with spatial multi-omics
Seminars in Cell and Developmental Biology,
Год журнала:
2025,
Номер
167, С. 10 - 21
Опубликована: Янв. 30, 2025
Язык: Английский
SpatialKNifeY (SKNY): Extending from spatial domain to surrounding area to identify microenvironment features with single-cell spatial omics data
PLoS Computational Biology,
Год журнала:
2025,
Номер
21(2), С. e1012854 - e1012854
Опубликована: Фев. 18, 2025
Single-cell
spatial
omics
analysis
requires
consideration
of
biological
functions
and
mechanisms
in
a
microenvironment.
However,
microenvironment
using
bioinformatic
methods
is
limited
by
the
need
to
detect
histological
morphology
extend
it
surrounding
area.
In
this
study,
we
developed
SpatialKNifeY
(SKNY),
an
image-processing-based
toolkit
that
detects
domains
potentially
reflect
histology
extends
these
Using
transcriptomic
data
from
breast
cancer,
applied
SKNY
algorithm
identify
tumor
domains,
followed
clustering
trajectory
estimation,
extension
(TME).
The
results
estimation
were
consistent
with
known
cancer
progression.
We
observed
vascularization
immunodeficiency
at
mid-
late-stage
progression
TME.
Furthermore,
integrate
cluster
14
patients
metastatic
colorectal
clusters
divided
based
on
TME
characteristics.
conclusion,
facilitates
determination
cataloguing
features.
Язык: Английский
spacedeconv: deconvolution of tissue architecture from spatial transcriptomics
Research Square (Research Square),
Год журнала:
2024,
Номер
unknown
Опубликована: Окт. 4, 2024
Abstract
Investigating
tissue
architecture
is
key
to
understanding
function
in
health
and
disease.
While
spatial
omics
technologies
enable
the
study
of
cell
transcriptomes
within
their
native
context,
they
often
lack
single-cell
resolution.
Deconvolution
methods
can
computationally
infer
composition
from
transcriptomics
data,
but
differences
workflows
complicate
use
comparison.
We
developed
spacedeconv,
a
unified
interface
different
deconvolution
that
additionally
supports
data
preprocessing,
visualization,
analysis
communication
multimodal
data.
Here,
we
demonstrate
how
spacedeconv
streamlines
investigation
cellular
molecular
underpinnings
organisms
contexts.
Язык: Английский
An integrative spatial multi-omic workflow for unified analysis of tumor tissue
bioRxiv (Cold Spring Harbor Laboratory),
Год журнала:
2024,
Номер
unknown
Опубликована: Окт. 18, 2024
Abstract
Combining
molecular
profiling
with
imaging
techniques
has
advanced
the
field
of
spatial
biology,
offering
new
insights
into
complex
biological
processes.
Focusing
on
diffuse
IDH
-mutated
low-grade
glioma,
this
study
presents
a
workflow
for
Spatial
Multi-omics
Integration,
SMINT,
specifically
combining
transcriptomics
and
metabolomics.
Our
incorporates
both
existing
custom-developed
computational
tools
to
enable
cell
segmentation
registration
coordinates
from
modalities
common
coordinate
framework.
During
our
investigation
strategies,
we
found
that
nuclei-only
segmentation,
while
containing
only
40%
segmented
transcripts,
enables
accurate
type
annotation,
but
does
not
account
multinucleated
cells.
integrative
including
cell-morphology
identified
distinct
cellular
neighborhoods
at
infiltrating
edge
gliomas,
which
were
enriched
in
oligodendrocyte-lineage
tumor
cells,
may
drive
invasion
normal
cortical
layers
brain.
Highlights
Alignment
integrated
analysis
transcriptomic
metabolomic
data
Nuclei-only
segmentations
are
concordant
annotation
Spatially
regions
conserved
datasets
Multi-omic
exploration
glioma
leading
identifies
novel
features
Язык: Английский
Deep learning pipeline for automated cell profiling from cyclic imaging
Scientific Reports,
Год журнала:
2024,
Номер
14(1)
Опубликована: Окт. 9, 2024
Cyclic
fluorescence
microscopy
enables
multiple
targets
to
be
detected
simultaneously.
This,
in
turn,
has
deepened
our
understanding
of
tissue
composition,
cell-to-cell
interactions,
and
cell
signaling.
Unfortunately,
analysis
these
datasets
can
time-prohibitive
due
the
sheer
volume
data.
In
this
paper,
we
present
CycloNET,
a
computational
pipeline
tailored
for
analyzing
raw
fluorescent
images
obtained
through
cyclic
immunofluorescence.
The
automated
pre-processes
image
files,
quickly
corrects
translation
errors
between
imaging
cycles,
leverages
pre-trained
neural
network
segment
individual
cells
generate
single-cell
molecular
profiles.
We
applied
CycloNET
dataset
22
human
samples
from
head
neck
squamous
carcinoma
patients
trained
immune
cells.
efficiently
processed
large-scale
(17
fields
view
per
cycle
13
staining
cycles
specimen)
10
min,
delivering
insights
at
resolution
facilitating
identification
rare
clusters.
expect
that
rapid
will
serve
as
powerful
tool
understand
complex
biological
systems
cellular
level,
with
potential
facilitate
breakthroughs
areas
such
developmental
biology,
disease
pathology,
personalized
medicine.
Язык: Английский
mxfda: A comprehensive toolkit for functional data analysis of single-cell spatial data
Bioinformatics Advances,
Год журнала:
2024,
Номер
4(1)
Опубликована: Янв. 1, 2024
Abstract
Summary
Technologies
that
produce
spatial
single-cell
(SC)
data
have
revolutionized
the
study
of
tissue
microstructures
and
promise
to
advance
personalized
treatment
cancer
by
revealing
new
insights
about
tumor
microenvironment.
Functional
analysis
(FDA)
is
an
ideal
analytic
framework
for
connecting
cell
relationships
patient
outcomes,
but
can
be
challenging
implement.
To
address
this
need,
we
present
mxfda,
R
package
end-to-end
SC
using
FDA.
mxfda
implements
a
suite
methods
facilitate
imaging
FDA
techniques.
Availability
implementation
The
freely
available
at
https://cran.r-project.org/package=mxfda
has
detailed
documentation,
including
four
vignettes,
http://juliawrobel.com/mxfda/.
Язык: Английский
MEGA-FISH: Multi-omics Extensible GPU-Accelerated FISH Processing Framework for Huge-Scale Spatial Omics
bioRxiv (Cold Spring Harbor Laboratory),
Год журнала:
2024,
Номер
unknown
Опубликована: Дек. 9, 2024
ABSTRACT
Spatial
omics
enables
comprehensive
mapping
of
cell
types
and
states
in
their
spatial
context,
providing
profound
insights
into
cellular
communication
tissue
organization.
However,
analyzing
large
sections,
especially
crucial
for
clinical
applications,
remains
a
significant
challenge
due
to
the
computational
demands
current
image
processing
methods.
To
overcome
these
limitations,
we
developed
MEGA-FISH,
flexible,
GPU-accelerated
Python
framework
optimized
large-scale
analysis.
Benchmarking
on
simulated
images
demonstrated
that
MEGA-FISH
achieved
high
accuracy
spot
detection
while
significantly
reducing
times
compared
with
established
tools.
The
framework’s
adaptable
capabilities
optimize
resource
allocation
(e.g.,
GPU
or
multi-core
CPU)
diverse
tasks,
its
scalable
architecture
integration
advanced
imaging
segmentation
techniques.
By
bridging
cutting-edge
methods
single-cell
analysis,
provides
an
efficient
platform
multi-modal
analysis
advances
research
applications
at
organ
organism
scales.
Язык: Английский