PLoS Computational Biology,
Journal Year:
2022,
Volume and Issue:
18(4), P. e1010066 - e1010066
Published: April 21, 2022
Machine
learning-based
classification
approaches
are
widely
used
to
predict
host
phenotypes
from
microbiome
data.
Classifiers
typically
employed
by
considering
operational
taxonomic
units
or
relative
abundance
profiles
as
input
features.
Such
types
of
data
intrinsically
sparse,
which
opens
the
opportunity
make
predictions
presence/absence
rather
than
microbial
taxa.
This
also
poses
question
whether
it
is
presence
particular
taxa
be
relevant
for
discrimination
purposes,
an
aspect
that
has
been
so
far
overlooked
in
literature.
In
this
paper,
we
aim
at
filling
gap
performing
a
meta-analysis
on
4,128
publicly
available
metagenomes
associated
with
multiple
case-control
studies.
At
species-level
resolution,
show
specific
important
when
building
models.
findings
robust
choice
classifier
and
confirmed
statistical
tests
applied
identifying
differentially
abundant/present
Results
further
coarser
resolutions
validated
4,026
additional
16S
rRNA
samples
coming
30
public
Frontiers in Microbiology,
Journal Year:
2021,
Volume and Issue:
12
Published: Feb. 19, 2021
The
number
of
microbiome-related
studies
has
notably
increased
the
availability
data
on
human
microbiome
composition
and
function.
These
provide
essential
material
to
deeply
explore
host-microbiome
associations
their
relation
development
progression
various
complex
diseases.
Improved
data-analytical
tools
are
needed
exploit
all
information
from
these
biological
datasets,
taking
into
account
peculiarities
data,
i.e.,
compositional,
heterogeneous
sparse
nature
datasets.
possibility
predicting
host-phenotypes
based
taxonomy-informed
feature
selection
establish
an
association
between
predict
disease
states
is
beneficial
for
personalized
medicine.
In
this
regard,
machine
learning
(ML)
provides
new
insights
models
that
can
be
used
outputs,
such
as
classification
prediction
in
microbiology,
infer
host
phenotypes
diseases
use
microbial
communities
stratify
patients
by
characterization
state-specific
signatures.
Here
we
review
state-of-the-art
ML
methods
respective
software
applied
studies,
performed
part
COST
Action
ML4Microbiome
activities.
This
scoping
focuses
application
related
clinical
diagnostics,
prognostics,
therapeutics.
Although
presented
here
more
bacterial
community,
many
algorithms
could
general,
regardless
type.
literature
covering
broad
topic
aligned
with
methodology.
manual
identification
sources
been
complemented
with:
(1)
automated
publication
search
through
digital
libraries
three
major
publishers
using
natural
language
processing
(NLP)
Toolkit,
(2)
relevant
repositories
GitHub
ranking
research
papers
relying
rank
approach.
The Journal of Physical Chemistry B,
Journal Year:
2022,
Volume and Issue:
126(34), P. 6372 - 6383
Published: Aug. 17, 2022
AlphaFold
has
burst
into
our
lives.
A
powerful
algorithm
that
underscores
the
strength
of
biological
sequence
data
and
artificial
intelligence
(AI).
appended
projects
research
directions.
The
database
it
been
creating
promises
an
untold
number
applications
with
vast
potential
impacts
are
still
difficult
to
surmise.
AI
approaches
can
revolutionize
personalized
treatments
usher
in
better-informed
clinical
trials.
They
promise
make
giant
leaps
toward
reshaping
revamping
drug
discovery
strategies,
selecting
prioritizing
combinations
targets.
Here,
we
briefly
overview
structural
biology,
including
molecular
dynamics
simulations
prediction
microbiota-human
protein-protein
interactions.
We
highlight
advancements
accomplished
by
deep-learning-powered
protein
structure
their
impact
on
life
sciences.
At
same
time,
does
not
resolve
decades-long
folding
challenge,
nor
identify
pathways.
models
provides
do
capture
conformational
mechanisms
like
frustration
allostery,
which
rooted
ensembles,
controlled
dynamic
distributions.
Allostery
signaling
properties
populations.
also
generate
ensembles
intrinsically
disordered
proteins
regions,
instead
describing
them
low
probabilities.
Since
generates
single
ranked
structures,
rather
than
cannot
elucidate
allosteric
activating
driver
hotspot
mutations
resistance.
However,
capturing
key
features,
deep
learning
techniques
use
predicted
conformation
as
basis
for
generating
a
diverse
ensemble.
World Journal of Gastroenterology,
Journal Year:
2022,
Volume and Issue:
28(4), P. 412 - 431
Published: Jan. 19, 2022
Irritable
bowel
syndrome
(IBS)
is
a
common
clinical
label
for
medically
unexplained
gastrointestinal
symptoms,
recently
described
as
disturbance
of
the
microbiota-gut-brain
axis.
Despite
decades
research,
pathophysiology
this
highly
heterogeneous
disorder
remains
elusive.
However,
dramatic
change
in
understanding
underlying
pathophysiological
mechanisms
surfaced
when
importance
gut
microbiota
protruded
scientific
picture.
Are
we
getting
any
closer
to
IBS'
etiology,
or
are
drowning
unspecific,
conflicting
data
because
possess
limited
tools
unravel
cluster
secrets
our
concealing?
In
comprehensive
review
discussing
some
major
important
features
IBS
and
their
interaction
with
microbiota,
microbiota-altering
treatment
such
low
FODMAP
diet
fecal
transplantation,
neuroimaging
methods
analyses,
current
future
challenges
big
analysis
IBS.
Bioinformatics,
Journal Year:
2023,
Volume and Issue:
39(2)
Published: Jan. 11, 2023
Machine
learning
(ML)
methods
are
motivated
by
the
need
to
automate
information
extraction
from
large
datasets
in
order
support
human
users
data-driven
tasks.
This
is
an
attractive
approach
for
integrative
joint
analysis
of
vast
amounts
omics
data
produced
next
generation
sequencing
and
other
-omics
assays.
A
systematic
assessment
current
literature
can
help
identify
key
trends
potential
gaps
methodology
applications.
We
surveyed
on
ML
multi-omic
integration
quantitatively
explored
goals,
techniques
involved
this
field.
were
particularly
interested
examining
how
researchers
use
deal
with
volume
complexity
these
datasets.Our
main
finding
that
used
those
address
challenges
few
samples
many
features.
Dimensionality
reduction
reduce
feature
count
alongside
models
also
appropriately
handle
relatively
samples.
Popular
include
autoencoders,
random
forests
vector
machines.
found
field
heavily
influenced
The
Cancer
Genome
Atlas
dataset,
which
accessible
contains
diverse
experiments.All
processing
scripts
available
at
GitLab
repository:
https://gitlab.com/polavieja_lab/ml_multi-omics_review/
or
Zenodo:
https://doi.org/10.5281/zenodo.7361807.Supplementary
Bioinformatics
online.
Frontiers in Microbiology,
Journal Year:
2023,
Volume and Issue:
14
Published: Sept. 22, 2023
Microbiome
data
predictive
analysis
within
a
machine
learning
(ML)
workflow
presents
numerous
domain-specific
challenges
involving
preprocessing,
feature
selection,
modeling,
performance
estimation,
model
interpretation,
and
the
extraction
of
biological
information
from
results.
To
assist
decision-making,
we
offer
set
recommendations
on
algorithm
pipeline
creation
evaluation,
stemming
COST
Action
ML4Microbiome.
We
compared
suggested
approaches
multi-cohort
shotgun
metagenomics
dataset
colorectal
cancer
patients,
focusing
their
in
disease
diagnosis
biomarker
discovery.
It
is
demonstrated
that
use
compositional
transformations
filtering
methods
as
part
preprocessing
does
not
always
improve
model.
In
contrast,
multivariate
such
Statistically
Equivalent
Signatures
algorithm,
was
effective
reducing
classification
error.
When
validated
separate
test
dataset,
this
combination
with
random
forest
provided
most
accurate
estimates.
Lastly,
showed
how
linear
modeling
by
logistic
regression
coupled
visualization
techniques
Individual
Conditional
Expectation
(ICE)
plots
can
yield
interpretable
results
insights.
These
findings
are
significant
for
clinicians
non-experts
alike
translational
applications.
Expert Review of Molecular Diagnostics,
Journal Year:
2024,
Volume and Issue:
24(3), P. 201 - 218
Published: Feb. 13, 2024
Introduction
Gut
microbes
pose
challenges
like
colon
inflammation,
deadly
diarrhea,
antimicrobial
resistance
dissemination,
and
chronic
disease
onset.
Development
of
early,
rapid
specific
diagnosis
tools
is
essential
for
improving
infection
control.
Point-of-care
testing
(POCT)
systems
offer
rapid,
sensitive,
low-cost
sample-to-answer
methods
microbe
detection
from
various
clinical
environmental
samples,
bringing
the
advantages
portability,
automation,
simple
operation.
Alzheimer s & Dementia,
Journal Year:
2023,
Volume and Issue:
19(11), P. 5209 - 5231
Published: June 7, 2023
Abstract
Microbial
infections
of
the
brain
can
lead
to
dementia,
and
for
many
decades
microbial
have
been
implicated
in
Alzheimer's
disease
(AD)
pathology.
However,
a
causal
role
infection
AD
remains
contentious,
lack
standardized
detection
methodologies
has
led
inconsistent
detection/identification
microbes
brains.
There
is
need
consensus
methodology;
Pathobiome
Initiative
aims
perform
comparative
molecular
analyses
post
mortem
brains
versus
cerebrospinal
fluid,
blood,
olfactory
neuroepithelium,
oral/nasopharyngeal
tissue,
bronchoalveolar,
urinary,
gut/stool
samples.
Diverse
extraction
methodologies,
polymerase
chain
reaction
sequencing
techniques,
bioinformatic
tools
will
be
evaluated,
addition
direct
culture
metabolomic
techniques.
The
goal
provide
roadmap
detecting
infectious
agents
patients
with
mild
cognitive
impairment
or
AD.
Positive
findings
would
then
prompt
tailoring
antimicrobial
treatments
that
might
attenuate
remit
mounting
clinical
deficits
subset
patients.