FEMS Microbiology Reviews,
Год журнала:
2024,
Номер
unknown
Опубликована: Дек. 17, 2024
Abstract
Microbial
functional
ecology
is
expanding
as
we
can
now
measure
the
traits
of
wild
microbes
that
affect
ecosystem
functioning.
Here,
review
techniques
and
advances
could
be
bedrock
for
a
unified
framework
to
study
microbial
functions.
These
include
our
newfound
access
environmental
genomes,
collections
traits,
but
also
ability
microbes’
distribution
expression.
We
then
explore
technical,
ecological,
evolutionary
processes
explain
patterns
diversity
redundancy.
Next,
suggest
reconciling
microbiology
with
biodiversity-ecosystem-functioning
studies
by
experimentally
testing
significance
redundancy
efficiency,
resistance,
resilience
processes.
Such
will
aid
in
identifying
state
shifts
tipping
points
microbiomes,
enhancing
understanding
how
where
microbiomes
guide
Earth's
biomes
context
changing
planet.
Cell Genomics,
Год журнала:
2022,
Номер
2(5), С. 100123 - 100123
Опубликована: Апрель 28, 2022
Marine
planktonic
eukaryotes
play
critical
roles
in
global
biogeochemical
cycles
and
climate.
However,
their
poor
representation
culture
collections
limits
our
understanding
of
the
evolutionary
history
genomic
underpinnings
ecosystems.
Here,
we
used
280
billion
Nature Biotechnology,
Год журнала:
2023,
Номер
42(6), С. 975 - 985
Опубликована: Сен. 7, 2023
Abstract
Exploiting
sequence–structure–function
relationships
in
biotechnology
requires
improved
methods
for
aligning
proteins
that
have
low
sequence
similarity
to
previously
annotated
proteins.
We
develop
two
deep
learning
address
this
gap,
TM-Vec
and
DeepBLAST.
allows
searching
structure–structure
similarities
large
databases.
It
is
trained
accurately
predict
TM-scores
as
a
metric
of
structural
directly
from
pairs
without
the
need
intermediate
computation
or
solution
structures.
Once
structurally
similar
been
identified,
DeepBLAST
can
align
using
only
information
by
identifying
homologous
regions
between
outperforms
traditional
alignment
performs
similarly
structure-based
methods.
show
merits
on
variety
datasets,
including
better
identification
remotely
compared
with
state-of-the-art
structure
prediction
Nature Communications,
Год журнала:
2024,
Номер
15(1)
Опубликована: Апрель 3, 2024
Abstract
Deciphering
the
relationship
between
a
gene
and
its
genomic
context
is
fundamental
to
understanding
engineering
biological
systems.
Machine
learning
has
shown
promise
in
latent
relationships
underlying
sequence-structure-function
paradigm
from
massive
protein
sequence
datasets.
However,
date,
limited
attempts
have
been
made
extending
this
continuum
include
higher
order
information.
Evolutionary
processes
dictate
specificity
of
contexts
which
found
across
phylogenetic
distances,
these
emergent
patterns
can
be
leveraged
uncover
functional
products.
Here,
we
train
language
model
(gLM)
on
millions
metagenomic
scaffolds
learn
regulatory
genes.
gLM
learns
contextualized
embeddings
that
capture
as
well
itself,
encode
biologically
meaningful
functionally
relevant
information
(e.g.
enzymatic
function,
taxonomy).
Our
analysis
attention
demonstrates
co-regulated
modules
(i.e.
operons).
findings
illustrate
gLM’s
unsupervised
deep
corpus
an
effective
promising
approach
semantics
syntax
genes
their
complex
region.
Nucleic Acids Research,
Год журнала:
2025,
Номер
53(3)
Опубликована: Янв. 24, 2025
Abstract
Many
universally
and
conditionally
important
genes
are
genomically
aggregated
within
clusters.
Here,
we
introduce
fai
zol,
which
together
enable
large-scale
comparative
analysis
of
different
types
gene
clusters
mobile-genetic
elements,
such
as
biosynthetic
(BGCs)
or
viruses.
Fundamentally,
they
overcome
a
current
bottleneck
to
reliably
perform
comprehensive
orthology
inference
at
large
scale
across
broad
taxonomic
contexts
thousands
genomes.
First,
allows
the
identification
orthologous
instances
query
cluster
interest
amongst
database
target
Subsequently,
zol
enables
reliable,
context-specific
ortholog
groups
for
individual
protein-encoding
instances.
In
addition,
performs
functional
annotation
computes
variety
evolutionary
statistics
each
inferred
group.
Importantly,
in
comparison
tools
visual
exploration
homologous
relationships
between
clusters,
can
handle
produce
detailed
reports
that
easy
digest.
To
showcase
apply
them
for:
(i)
longitudinal
tracking
virus
metagenomes,
(ii)
performing
population
genetic
investigations
BGCs
fungal
species,
(iii)
uncovering
trends
virulence-associated
genomes
from
diverse
bacterial
genus.
Nature,
Год журнала:
2023,
Номер
626(7998), С. 377 - 384
Опубликована: Дек. 18, 2023
Many
of
the
Earth's
microbes
remain
uncultured
and
understudied,
limiting
our
understanding
functional
evolutionary
aspects
their
genetic
material,
which
largely
overlooked
in
most
metagenomic
studies
Nucleic Acids Research,
Год журнала:
2023,
Номер
52(D1), С. D777 - D783
Опубликована: Окт. 28, 2023
Abstract
Meta’omic
data
on
microbial
diversity
and
function
accrue
exponentially
in
public
repositories,
but
derived
information
is
often
siloed
according
to
type,
study
or
sampled
environment.
Here
we
present
SPIRE,
a
Searchable
Planetary-scale
mIcrobiome
REsource
that
integrates
various
consistently
processed
metagenome-derived
modalities
across
habitats,
geography
phylogeny.
SPIRE
encompasses
99
146
metagenomic
samples
from
739
studies
covering
wide
array
of
environments
augmented
with
manually-curated
contextual
data.
Across
total
assembly
16
Tbp,
comprises
35
billion
predicted
protein
sequences
1.16
million
newly
constructed
metagenome-assembled
genomes
(MAGs)
medium
high
quality.
Beyond
mapping
the
high-quality
genome
reference
provided
by
proGenomes3
(http://progenomes.embl.de),
these
novel
MAGs
form
92
134
species-level
clusters,
majority
which
are
unclassified
at
species
level
using
current
tools.
enables
taxonomic
profiling
clusters
via
an
updated,
custom
mOTUs
database
(https://motu-tool.org/)
includes
several
layers
functional
annotation,
as
well
crosslinks
(micro-)biological
databases.
The
resource
accessible,
searchable
browsable
http://spire.embl.de.
