<s>
Machine	B-Application
learning	I-Application
in	I-Application
bioinformatics	I-Application
is	O
the	O
application	O
of	O
machine	O
learning	O
algorithms	O
to	O
bioinformatics	O
,	O
including	O
genomics	O
,	O
proteomics	O
,	O
microarrays	O
,	O
systems	O
biology	O
,	O
evolution	O
,	O
and	O
text	B-Algorithm
mining	I-Algorithm
.	O
</s>
<s>
Machine	O
learning	O
techniques	O
,	O
such	O
as	O
deep	B-Algorithm
learning	I-Algorithm
can	O
learn	B-General_Concept
features	I-General_Concept
of	O
data	O
sets	O
,	O
instead	O
of	O
requiring	O
the	O
programmer	O
to	O
define	O
them	O
individually	O
.	O
</s>
<s>
The	O
algorithm	O
can	O
further	O
learn	O
how	O
to	O
combine	O
low-level	O
features	B-Algorithm
into	O
more	O
abstract	O
features	B-Algorithm
,	O
and	O
so	O
on	O
.	O
</s>
<s>
Machine	O
learning	O
algorithms	O
in	O
bioinformatics	O
can	O
be	O
used	O
for	O
prediction	O
,	O
classification	O
,	O
and	O
feature	B-General_Concept
selection	I-General_Concept
.	O
</s>
<s>
The	O
type	O
of	O
algorithm	O
,	O
or	O
process	O
used	O
to	O
build	O
the	O
predictive	O
models	O
from	O
data	O
using	O
analogies	O
,	O
rules	O
,	O
neural	B-Architecture
networks	I-Architecture
,	O
probabilities	O
,	O
and/or	O
statistics	O
.	O
</s>
<s>
Artificial	B-Architecture
neural	I-Architecture
networks	I-Architecture
in	O
bioinformatics	O
have	O
been	O
used	O
for	O
:	O
</s>
<s>
The	O
way	O
that	O
features	B-Algorithm
,	O
often	O
vectors	O
in	O
a	O
many-dimensional	O
space	O
,	O
are	O
extracted	O
from	O
the	O
domain	O
data	O
is	O
an	O
important	O
component	O
of	O
learning	O
systems	O
.	O
</s>
<s>
in	O
this	O
case	O
the	O
dimension	O
is	O
)	O
,	O
techniques	O
such	O
as	O
principal	B-Application
component	I-Application
analysis	I-Application
are	O
used	O
to	O
project	O
the	O
data	O
to	O
a	O
lower	O
dimensional	O
space	O
,	O
thus	O
selecting	O
a	O
smaller	O
set	O
of	O
features	B-Algorithm
from	O
the	O
sequences	O
.	O
</s>
<s>
Convolutional	B-Architecture
neural	I-Architecture
networks	I-Architecture
(	O
CNN	B-Architecture
)	O
are	O
a	O
class	O
of	O
deep	O
neural	B-Architecture
network	I-Architecture
whose	O
architecture	O
is	O
based	O
on	O
shared	O
weights	O
of	O
convolution	O
kernels	O
or	O
filters	O
that	O
slide	O
along	O
input	O
features	B-Algorithm
,	O
providing	O
translation-equivariant	O
responses	O
known	O
as	O
feature	O
maps	O
.	O
</s>
<s>
CNNs	B-Architecture
take	O
advantage	O
of	O
the	O
hierarchical	O
pattern	O
in	O
data	O
and	O
assemble	O
patterns	O
of	O
increasing	O
complexity	O
using	O
smaller	O
and	O
simpler	O
patterns	O
discovered	O
via	O
their	O
filters	O
.	O
</s>
<s>
Convolutional	O
networks	O
were	O
inspired	O
by	O
biological	O
processes	O
in	O
that	O
the	O
connectivity	O
pattern	O
between	O
neurons	B-Algorithm
resembles	O
the	O
organization	O
of	O
the	O
animal	O
visual	O
cortex	O
.	O
</s>
<s>
Individual	O
cortical	O
neurons	B-Algorithm
respond	O
to	O
stimuli	O
only	O
in	O
a	O
restricted	O
region	O
of	O
the	O
visual	O
field	O
known	O
as	O
the	O
receptive	O
field	O
.	O
</s>
<s>
The	O
receptive	O
fields	O
of	O
different	O
neurons	B-Algorithm
partially	O
overlap	O
such	O
that	O
they	O
cover	O
the	O
entire	O
visual	O
field	O
.	O
</s>
<s>
CNN	B-Architecture
uses	O
relatively	O
little	O
pre-processing	O
compared	O
to	O
other	O
image	O
classification	O
algorithms	O
.	O
</s>
<s>
This	O
means	O
that	O
the	O
network	O
learns	O
to	O
optimize	O
the	O
filters	O
(	O
or	O
kernels	O
)	O
through	O
automated	O
learning	O
,	O
whereas	O
in	O
traditional	O
algorithms	O
these	O
filters	O
are	O
hand-engineered	B-General_Concept
.	O
</s>
<s>
This	O
independence	O
from	O
prior	O
knowledge	O
and	O
human	O
intervention	O
in	O
feature	B-Algorithm
extraction	I-Algorithm
is	O
a	O
major	O
advantage	O
.	O
</s>
<s>
A	O
phylogenetic	O
convolutional	B-Architecture
neural	I-Architecture
network	I-Architecture
(	O
Ph-CNN	O
)	O
is	O
a	O
novel	O
convolutional	B-Architecture
neural	I-Architecture
network	I-Architecture
architecture	O
proposed	O
by	O
Fioranti	O
et	O
al	O
.	O
</s>
<s>
Ph-CNN	O
achieves	O
promising	O
results	O
compared	O
to	O
fully	O
connected	O
neural	B-Architecture
networks	I-Architecture
,	O
random	B-Algorithm
forest	I-Algorithm
and	O
support	B-Algorithm
vector	I-Algorithm
machines	I-Algorithm
.	O
</s>
<s>
Random	B-Algorithm
forests	I-Algorithm
(	O
RF	O
)	O
classify	O
by	O
constructing	O
an	O
ensemble	O
of	O
decision	B-Algorithm
trees	I-Algorithm
,	O
and	O
outputting	O
the	O
average	O
prediction	O
of	O
the	O
individual	O
trees	O
.	O
</s>
<s>
This	O
is	O
a	O
modification	O
of	O
bootstrap	B-Algorithm
aggregating	I-Algorithm
(	O
which	O
aggregates	O
a	O
large	O
collection	O
of	O
decision	B-Algorithm
trees	I-Algorithm
)	O
and	O
can	O
be	O
used	O
for	O
classification	O
or	O
regression	O
.	O
</s>
<s>
As	O
random	B-Algorithm
forests	I-Algorithm
give	O
an	O
internal	O
estimate	O
of	O
generalization	O
error	O
,	O
cross-validation	O
is	O
unnecessary	O
.	O
</s>
<s>
Computationally	O
,	O
random	B-Algorithm
forests	I-Algorithm
are	O
appealing	O
because	O
they	O
naturally	O
handle	O
both	O
regression	O
and	O
(	O
multiclass	O
)	O
classification	O
,	O
are	O
relatively	O
fast	O
to	O
train	O
and	O
to	O
predict	O
,	O
depend	O
only	O
on	O
one	O
or	O
two	O
tuning	O
parameters	O
,	O
have	O
a	O
built-in	O
estimate	O
of	O
the	O
generalization	O
error	O
,	O
can	O
be	O
used	O
directly	O
for	O
high-dimensional	O
problems	O
,	O
and	O
can	O
easily	O
be	O
implemented	O
in	O
parallel	O
.	O
</s>
<s>
Statistically	O
,	O
random	B-Algorithm
forests	I-Algorithm
are	O
appealing	O
for	O
additional	O
features	B-Algorithm
,	O
such	O
as	O
measures	O
of	O
variable	O
importance	O
,	O
differential	O
class	O
weighting	O
,	O
missing	O
value	O
imputation	O
,	O
visualization	O
,	O
outlier	O
detection	O
,	O
and	O
unsupervised	O
learning	O
.	O
</s>
<s>
Clustering	B-Algorithm
-	O
the	O
partitioning	O
of	O
a	O
data	O
set	O
into	O
disjoint	O
subsets	O
,	O
so	O
that	O
the	O
data	O
in	O
each	O
subset	O
are	O
as	O
close	O
as	O
possible	O
to	O
each	O
other	O
and	O
as	O
distant	O
as	O
possible	O
from	O
data	O
in	O
any	O
other	O
subset	O
,	O
according	O
to	O
some	O
defined	O
distance	O
or	O
similarity	O
function	O
-	O
is	O
a	O
common	O
technique	O
for	O
statistical	O
data	O
analysis	O
.	O
</s>
<s>
Clustering	B-Algorithm
is	O
central	O
to	O
much	O
data-driven	O
bioinformatics	O
research	O
and	O
serves	O
as	O
a	O
powerful	O
computational	O
method	O
whereby	O
means	O
of	O
hierarchical	O
,	O
centroid-based	O
,	O
distribution-based	O
,	O
density-based	O
,	O
and	O
self-organizing	O
maps	O
classification	O
,	O
has	O
long	O
been	O
studied	O
and	O
used	O
in	O
classical	O
machine	O
learning	O
settings	O
.	O
</s>
<s>
Particularly	O
,	O
clustering	B-Algorithm
helps	O
to	O
analyze	O
unstructured	O
and	O
high-dimensional	O
data	O
in	O
the	O
form	O
of	O
sequences	O
,	O
expressions	O
,	O
texts	O
,	O
images	O
,	O
and	O
so	O
on	O
.	O
</s>
<s>
Clustering	B-Algorithm
is	O
also	O
used	O
to	O
gain	O
insights	O
into	O
biological	O
processes	O
at	O
the	O
genomic	O
level	O
,	O
e.g.	O
</s>
<s>
Data	B-Algorithm
clustering	I-Algorithm
algorithms	O
can	O
be	O
hierarchical	O
or	O
partitional	O
.	O
</s>
<s>
Hierarchical	B-Algorithm
clustering	I-Algorithm
is	O
calculated	O
using	O
metrics	O
on	O
Euclidean	O
spaces	O
,	O
the	O
most	O
commonly	O
used	O
is	O
the	O
Euclidean	O
distance	O
computed	O
by	O
finding	O
the	O
square	O
of	O
the	O
difference	O
between	O
each	O
variable	O
,	O
adding	O
all	O
the	O
squares	O
,	O
and	O
finding	O
the	O
square	O
root	O
of	O
the	O
said	O
sum	O
.	O
</s>
<s>
An	O
example	O
of	O
a	O
hierarchical	B-Algorithm
clustering	I-Algorithm
algorithm	O
is	O
BIRCH	B-Algorithm
,	O
which	O
is	O
particularly	O
good	O
on	O
bioinformatics	O
for	O
its	O
nearly	O
linear	O
time	O
complexity	O
given	O
generally	O
large	O
datasets	O
.	O
</s>
<s>
Most	O
applications	O
adopt	O
one	O
of	O
two	O
popular	O
heuristic	O
methods	O
:	O
k-means	B-Algorithm
algorithm	I-Algorithm
or	O
k-medoids	B-Algorithm
.	O
</s>
<s>
Other	O
algorithms	O
do	O
not	O
require	O
an	O
initial	O
number	O
of	O
groups	O
,	O
such	O
as	O
affinity	B-Algorithm
propagation	I-Algorithm
.	O
</s>
<s>
For	O
example	O
,	O
machine	O
learning	O
methods	O
can	O
be	O
trained	O
to	O
identify	O
specific	O
visual	O
features	B-Algorithm
such	O
as	O
splice	O
sites	O
.	O
</s>
<s>
Support	B-Algorithm
vector	I-Algorithm
machines	I-Algorithm
have	O
been	O
extensively	O
used	O
in	O
cancer	O
genomic	O
studies	O
.	O
</s>
<s>
In	O
addition	O
,	O
deep	B-Algorithm
learning	I-Algorithm
has	O
been	O
incorporated	O
into	O
bioinformatic	O
algorithms	O
.	O
</s>
<s>
Deep	B-Algorithm
learning	I-Algorithm
applications	O
have	O
been	O
used	O
for	O
regulatory	O
genomics	O
and	O
cellular	O
imaging	O
.	O
</s>
<s>
Deep	B-Algorithm
learning	I-Algorithm
has	O
been	O
applied	O
to	O
regulatory	O
genomics	O
,	O
variant	O
calling	O
and	O
pathogenicity	O
scores	O
.	O
</s>
<s>
Natural	B-Language
language	I-Language
processing	I-Language
and	O
text	B-Algorithm
mining	I-Algorithm
have	O
helped	O
to	O
understand	O
phenomena	O
including	O
protein-protein	O
interaction	O
,	O
gene-disease	O
relation	O
as	O
well	O
as	O
predicting	O
biomolecule	O
structures	O
and	O
functions	O
.	O
</s>
<s>
Natural	B-Language
language	I-Language
processing	I-Language
algorithms	O
personalized	O
medicine	O
for	O
patients	O
who	O
suffer	O
genetic	O
diseases	O
,	O
by	O
combining	O
the	O
extraction	O
of	O
clinical	O
information	O
and	O
genomic	O
data	O
available	O
from	O
the	O
patients	O
.	O
</s>
<s>
Automatic	O
feature	B-General_Concept
learning	I-General_Concept
reaches	O
an	O
accuracy	O
of	O
82-84	O
%	O
.	O
</s>
<s>
The	O
current	O
state-of-the-art	O
in	O
secondary	O
structure	O
prediction	O
uses	O
a	O
system	O
called	O
DeepCNF	O
(	O
deep	O
convolutional	O
neural	O
fields	O
)	O
which	O
relies	O
on	O
the	O
machine	O
learning	O
model	O
of	O
artificial	B-Architecture
neural	I-Architecture
networks	I-Architecture
to	O
achieve	O
an	O
accuracy	O
of	O
approximately	O
84%	O
when	O
tasked	O
to	O
classify	O
the	O
amino	O
acids	O
of	O
a	O
protein	O
sequence	O
into	O
one	O
of	O
three	O
structural	O
classes	O
(	O
helix	O
,	O
sheet	O
,	O
or	O
coil	O
)	O
.	O
</s>
<s>
16S	O
rRNA	O
or	O
whole-genome	O
sequencing	O
(	O
WGS	O
)	O
,	O
using	O
methods	O
such	O
as	O
least	O
absolute	O
shrinkage	O
and	O
selection	O
operator	O
classifier	O
,	O
random	B-Algorithm
forest	I-Algorithm
,	O
supervised	O
classification	O
model	O
,	O
and	O
gradient	O
boosted	O
tree	O
model	O
.	O
</s>
<s>
Neural	B-Architecture
networks	I-Architecture
,	O
such	O
as	O
recurrent	B-Algorithm
neural	I-Algorithm
networks	I-Algorithm
(	O
RNN	O
)	O
,	O
convolutional	B-Architecture
neural	I-Architecture
networks	I-Architecture
(	O
CNN	B-Architecture
)	O
,	O
and	O
Hopfield	B-Algorithm
neural	I-Algorithm
networks	I-Algorithm
have	O
been	O
added	O
.	O
</s>
<s>
developed	O
an	O
algorithm	O
called	O
Ph-CNN	O
to	O
classify	O
data	O
samples	O
from	O
healthy	O
patients	O
and	O
patients	O
with	O
IBD	O
symptoms	O
(	O
to	O
distinguish	O
healthy	O
and	O
sick	O
patients	O
)	O
by	O
using	O
phylogenetic	O
trees	O
and	O
convolutional	B-Architecture
neural	I-Architecture
networks	I-Architecture
.	O
</s>
<s>
In	O
addition	O
,	O
random	B-Algorithm
forest	I-Algorithm
(	O
RF	O
)	O
methods	O
and	O
implemented	O
importance	O
measures	O
help	O
in	O
the	O
identification	O
of	O
microbiome	O
species	O
that	O
can	O
be	O
used	O
to	O
distinguish	O
diseased	O
and	O
non-diseased	O
samples	O
.	O
</s>
<s>
However	O
,	O
the	O
performance	O
of	O
a	O
decision	B-Algorithm
tree	I-Algorithm
and	O
the	O
diversity	O
of	O
decision	B-Algorithm
trees	I-Algorithm
in	O
the	O
ensemble	O
significantly	O
influence	O
the	O
performance	O
of	O
RF	O
algorithms	O
.	O
</s>
<s>
Effective	O
approaches	O
require	O
many	O
possible	O
variable	O
combinations	O
,	O
which	O
exponentially	O
increases	O
the	O
computational	O
burden	O
as	O
the	O
number	O
of	O
features	B-Algorithm
increases	O
.	O
</s>
<s>
The	O
core	O
of	O
the	O
pipeline	O
is	O
an	O
RF	O
classifier	O
coupled	O
with	O
forwarding	O
variable	B-General_Concept
selection	I-General_Concept
(	O
RF-FVS	O
)	O
,	O
which	O
selects	O
a	O
minimum-size	O
core	O
set	O
of	O
microbial	O
species	O
or	O
functional	O
signatures	O
that	O
maximize	O
the	O
predictive	O
classifier	O
performance	O
.	O
</s>
<s>
The	O
most	O
commonly	O
used	O
methods	O
are	O
radial	B-Algorithm
basis	I-Algorithm
function	I-Algorithm
networks	I-Algorithm
,	O
deep	B-Algorithm
learning	I-Algorithm
,	O
Bayesian	B-General_Concept
classification	I-General_Concept
,	O
decision	B-Algorithm
trees	I-Algorithm
,	O
and	O
random	B-Algorithm
forest	I-Algorithm
.	O
</s>
<s>
Genetic	B-Algorithm
algorithms	I-Algorithm
,	O
machine	O
learning	O
techniques	O
which	O
are	O
based	O
on	O
the	O
natural	O
process	O
of	O
evolution	O
,	O
have	O
been	O
used	O
to	O
model	O
genetic	O
networks	O
and	O
regulatory	O
structures	O
.	O
</s>
<s>
This	O
domain	O
,	O
particularly	O
phylogenetic	O
tree	O
reconstruction	O
,	O
uses	O
the	O
features	B-Algorithm
of	O
machine	O
learning	O
techniques	O
.	O
</s>
<s>
Initially	O
,	O
they	O
were	O
constructed	O
using	O
features	B-Algorithm
such	O
as	O
morphological	O
and	O
metabolic	O
features	B-Algorithm
.	O
</s>
<s>
Machine	O
learning	O
methods	O
for	O
the	O
analysis	O
of	O
neuroimaging	B-Algorithm
data	O
are	O
used	O
to	O
help	O
diagnose	O
stroke	O
.	O
</s>
<s>
Historically	O
multiple	O
approaches	O
to	O
this	O
problem	O
involved	O
neural	B-Architecture
networks	I-Architecture
.	O
</s>
<s>
As	O
proposed	O
by	O
Titano	O
3D-CNN	O
techniques	O
were	O
tested	O
in	O
supervised	O
classification	O
to	O
screen	O
head	O
CT	O
images	O
for	O
acute	O
neurologic	O
events	O
.	O
</s>
<s>
Three-dimensional	O
CNN	B-Architecture
and	O
SVM	B-Algorithm
methods	O
are	O
often	O
used	O
.	O
</s>
<s>
Machine	O
learning	O
can	O
be	O
used	O
for	O
this	O
knowledge	O
extraction	O
task	O
using	O
techniques	O
such	O
as	O
natural	B-Language
language	I-Language
processing	I-Language
to	O
extract	O
the	O
useful	O
information	O
from	O
human-generated	O
reports	O
in	O
a	O
database	O
.	O
</s>
<s>
Text	B-General_Concept
Nailing	I-General_Concept
,	O
an	O
alternative	O
approach	O
to	O
machine	O
learning	O
,	O
capable	O
of	O
extracting	O
features	B-Algorithm
from	O
clinical	O
narrative	O
notes	O
was	O
introduced	O
in	O
2017	O
.	O
</s>
<s>
Another	O
application	O
of	O
text	B-Algorithm
mining	I-Algorithm
is	O
the	O
detection	O
and	O
visualization	O
of	O
distinct	O
DNA	O
regions	O
given	O
sufficient	O
reference	O
data	O
.	O
</s>
<s>
BiG-SLiCE	O
(	O
Biosynthetic	O
Genes	O
Super-Linear	O
Clustering	B-Algorithm
Engine	O
)	O
,	O
is	O
an	O
automated	O
pipeline	O
tool	O
designed	O
to	O
cluster	O
massive	O
numbers	O
of	O
BGCs	O
.	O
</s>
<s>
The	O
BiG-SLiCE	O
workflow	O
starts	O
at	O
vectorization	B-Algorithm
(	O
feature	B-Algorithm
extraction	I-Algorithm
)	O
,	O
converting	O
input	O
BGCs	O
provided	O
from	O
dataset	O
of	O
cluster	O
GenBank	O
files	O
from	O
antiSMASH	O
and	O
MIBiG	O
into	O
vectors	O
of	O
numerical	O
features	B-Algorithm
based	O
on	O
the	O
absence/presence	O
and	O
bitscores	O
of	O
hits	O
obtained	O
from	O
querying	O
BGC	O
gene	O
sequences	O
from	O
a	O
library	O
curated	O
of	O
profile	O
Hidden	O
Markov	O
Model(pHMMs )	O
of	O
biosynthetic	O
domains	O
of	O
BGCs	O
.	O
</s>
<s>
Those	O
features	B-Algorithm
are	O
then	O
processed	O
by	O
a	O
super-linear	O
clustering	B-Algorithm
algorithm	I-Algorithm
based	O
on	O
BIRCH	B-Algorithm
clustering	B-Algorithm
,	O
resulting	O
in	O
centroid	O
feature	B-Algorithm
vectors	I-Algorithm
representing	O
the	O
GCF	O
models	O
.	O
</s>
<s>
Then	O
a	O
global	O
cluster	O
mapping	O
is	O
done	O
using	O
k-means	B-Algorithm
to	O
group	O
all	O
GCF	O
centroid	O
features	B-Algorithm
in	O
GCF	O
bins	O
.	O
</s>
<s>
After	O
that	O
another	O
round	O
of	O
membership	O
assignment	O
is	O
performed	O
to	O
match	O
the	O
full	O
set	O
of	O
BGC	O
features	B-Algorithm
into	O
the	O
resulting	O
GCF	O
bins	O
.	O
</s>
<s>
Using	O
a	O
MinHash-based	O
algorithm	O
,	O
MASH	O
,	O
BiG-MAP	O
estimates	O
distance	O
among	O
protein	O
sequences	O
which	O
then	O
is	O
used	O
to	O
select	O
a	O
representative	O
gene	O
cluster	O
with	O
the	O
aid	O
of	O
k-medoids	B-Algorithm
clustering	I-Algorithm
.	O
</s>
<s>
Differential	O
expressions	O
analyses	O
use	O
zero-inflated	O
Gaussian	O
distribution	O
mixture	O
models	O
(	O
ZIG-models	O
)	O
or	O
Kruskal-Wallis	B-General_Concept
.	O
</s>
<s>
RiPPMiner	O
differentiates	O
RiPPs	O
from	O
other	O
proteins	O
and	O
peptides	O
using	O
a	O
support-vector	B-Algorithm
machine	I-Algorithm
model	O
trained	O
on	O
293	O
experimentally	O
characterized	O
RiPPs	O
as	O
a	O
positive	O
data	O
set	O
,	O
and	O
8140	O
genomes	O
encoded	O
non-RiPPs	O
polypeptides	O
as	O
negative	O
data	O
set	O
.	O
</s>
<s>
For	O
prediction	O
of	O
RiPP	O
class	O
or	O
sub-class	O
,	O
a	O
Multi-Class	O
SVM	B-Algorithm
was	O
trained	O
using	O
the	O
amino	O
acid	O
composition	O
and	O
dipeptide	O
frequencies	O
as	O
feature	B-Algorithm
vectors	I-Algorithm
.	O
</s>
<s>
During	O
the	O
training	O
of	O
the	O
Multi-Class	O
SVM	B-Algorithm
,	O
available	O
RiPP	O
precursor	O
sequences	O
belonging	O
to	O
a	O
given	O
class	O
(	O
e.g.	O
</s>
<s>
Out	O
of	O
the	O
four	O
major	O
RiPP	O
classes	O
that	O
had	O
more	O
than	O
50	O
experimentally	O
characterized	O
RiPPs	O
in	O
RiPPDB	O
,	O
SVM	B-Algorithm
models	O
for	O
prediction	O
of	O
cleavage	O
sites	O
were	O
developed	O
for	O
lanthipeptides	O
,	O
cyanobactins	O
,	O
and	O
lasso	O
peptides	O
.	O
</s>
<s>
In	O
order	O
to	O
develop	O
SVM	B-Algorithm
for	O
prediction	O
of	O
cleavage	O
site	O
for	O
lanthipeptides	O
,	O
12	O
mer	O
peptide	O
sequences	O
centered	O
on	O
the	O
cleavage	O
sites	O
were	O
extracted	O
from	O
a	O
set	O
of	O
115	O
lanthipeptide	O
precursor	O
sequences	O
with	O
known	O
cleavage	O
pattern	O
.	O
</s>
<s>
Feature	B-Algorithm
vectors	I-Algorithm
for	O
each	O
of	O
these	O
mers	O
consisted	O
of	O
the	O
concatenation	O
of	O
20-dimensional	O
vectors	O
corresponding	O
to	O
each	O
of	O
the	O
20	O
amino	O
acids	O
.	O
</s>
<s>
An	O
SVM	B-Algorithm
model	O
for	O
prediction	O
of	O
cleavage	O
site	O
was	O
developed	O
and	O
benchmarked	O
using	O
2-fold	O
cross-validation	O
,	O
where	O
half	O
of	O
the	O
data	O
were	O
used	O
in	O
training	O
and	O
the	O
other	O
half	O
in	O
testing	O
.	O
</s>
<s>
SVM	B-Algorithm
models	O
were	O
developed	O
for	O
the	O
prediction	O
of	O
the	O
cleavage	O
sites	O
in	O
cyanobactin	O
and	O
lasso	O
peptides	O
.	O
</s>
<s>
Spec2vec	O
algorithm	O
provides	O
a	O
new	O
way	O
of	O
spectral	O
similarity	O
score	O
,	O
based	O
on	O
Word2Vec	B-Algorithm
.	O
</s>
<s>
The	O
National	O
Center	O
for	O
Biotechnology	O
Information	O
(	O
NCBI	O
)	O
provides	O
a	O
large	O
suite	O
of	O
online	O
resources	O
for	O
biological	O
information	O
and	O
data	O
,	O
including	O
the	O
GenBank	O
nucleic	O
acid	O
sequence	O
database	O
and	O
the	O
PubMed	B-Library
database	O
of	O
citations	O
and	O
abstracts	O
for	O
published	O
life	O
science	O
journals	O
.	O
</s>
<s>
Resources	O
include	O
PubMed	B-Library
Data	O
Management	O
,	O
RefSeq	O
Functional	O
Elements	O
,	O
genome	O
data	O
download	O
,	O
variation	O
services	O
API	O
,	O
Magic-BLAST	O
,	O
QuickBLASTp	O
,	O
and	O
Identical	O
Protein	O
Groups	O
.	O
</s>
