PDBefold Tutorial
PDBefold Tutorial
http://pdbe.org
PDBeFOLD Tutorial
PDBeFOLD
may
be
accessed
from
multiple
locations
on
the
PDBe
website.
From
the
PDBe
home
page
(http://pdbe.org/),
there
are
two
access
points
for
the
program
as
shown
below.
The
link
highlighted
in
blue
may
be
used
to
automatically
get
all
related
structures
to
a
given
PDB
code.
In
order
to
start
the
webservice,
click
on
the
link
shown
in
red
above.
This
will
open
up
a
introductory
page
concerning
PDBeFOLD
that
contains
additional
information,
tips
and
help
regarding
the
webservice.
On
this
page,
now
click
on
the
Start
PDBeFOLD
button.
This
should
start
up
the
actual
webservice
that
will
in
the
form
of
a
submission
form
as
shown
below.
The
form
contains
parameters
that
may
be
adjusted
or
tweaked
in
order
to
customize
the
results
as
required.
The
Query
contains
information
regarding
the
structure/structures
that
need
to
be
compared.
This
could
be
an
existing
PDB
http://pdbe.org
PDBeFOLD Tutorial
code,
an
uploaded
coordinate
file,
SCOP
entry
or
pair
of
existing
PDB
entries
that
need
to
be
compared.
In
addition
to
this,
if
an
entry
contains
more
than
one
chain,
the
chain
that
has
to
be
compared
can
be
chosen.
Similarly
the
target
for
the
search
could
be
another
PDB
entry,
uploaded
coordinate
file
or
SCOP
set
or
a
set
of
files.
The
lowest
acceptable
match
boxes
tell
the
program
what
cut-off
to
use
in
the
matching
process.
The
default
is
70%
both
for
the
query
and
target,
meaning
that
in
order
to
list
a
comparison
as
a
match,
at
least
70%
of
the
secondary
structure
of
the
query
must
match
70%
of
the
target
structure.
These
settings
can
be
adjusted.
Setting
this
to
a
lower
value
will
result
in
a
large
number
of
hits
and
the
opposite
may
result
in
only
identical
structures
found.
We
will
use
the
PDB
entry
2MJP
(http://pdbe.org/2mjp)
(STRUCTURE-BASED
IDENTIFICATION
OF
THE
BIOCHEMICAL
FUNCTION
OF
A
HYPOTHETICAL
PROTEIN
FROM
METHANOCOCCUS
JANNASCHII:MJ0226)
as
an
example
for
demonstrating
PDBefold.
Click
on
the
Submit
your
query
button.
This
will
start
a
alignment
job
on
the
farm
which
will
take
upto
2-3
minutes
depending
on
server
load.
Once
the
results
are
calculated,
a
new
page
showing
the
same
will
be
displayed
on
the
browser
(Similar
to
that
shown
below).
http://pdbe.org
PDBeFOLD Tutorial
The
results
here
are
sorted
based
on
Q-score
(Quality
of
alignment,
with
1
being
the
highest
score)
for
this
entry.
The
pairwise
alignment
result
between
2mjp
and
2e5x
is
highlighted
in
the
above
figure.
There
is
a
49%
amino
acid
sequence
identity
between
the
two
proteins,
whereas
they
have
93%
secondary
structure
identity.
Clicking
on
the
number
link
on
the
left
hand
side
of
the
page
will
return
residue-by-residue
description
about
the
structural
alignment
between
the
two
proteins.
Note:
as
the
PDB
archive
continuously
adds
new
entries
every
week,
it
is
possible
that
this
result
may
not
appear
in
the
same
position
as
shown.
http://pdbe.org
PDBeFOLD Tutorial
Click
on
the
View
Superposed
button
to
view
the
secondary
structure
alignment.
The
two
structures
2MJP
and
2E5X
appear
share
a
high
level
of
secondary
structure
identity
despite
only
having
<50%
sequence
identity.
This
indicates
that
the
two
proteins
belong
to
the
same
structural
family.
A
little
further
down
the
page
are
the
summary
of
the
alignment
showing
the
secondary
structure
elements
that
match
between
the
two
structures
with
residue
ranges
and
the
matrix
required
to
move
and
align
the
target
to
the
query
structure.
http://pdbe.org
PDBeFOLD Tutorial
If
you
scroll
down
the
page,
there
is
a
3D
structural
alignment
between
the
residues
from
the
corresponding
PDB
entries.
This
alignment
shows
the
extent
of
superposition
between
similar
secondary
structure
folds.
The
red
background
indicates
that
matched
residues
have
same
residue
name,
cyan
background
is
used
for
other
matched
residues.
Unmatched
residues
are
indicated
by
black
background.
The
letters
H
and
S
represent
helix
and
-strands
respectively.
http://pdbe.org
PDBeFOLD Tutorial
The
residue-by-residue
mapping
provides
a
very
useful
tool
to
analyse
the
structure
function
relationship
between
the
two
entries.
Lets
now
concentrate
on
the
proteins
themselves
to
get
more
information
about
them.
You
can
go
to
the
summary
pages
for
these
entries
at
the
PDBe
in
new
browser
windows
or
tabs.
Go
to
http://www.pdbe.org/2mjp
in
one
window
and
http://www.pdbe.org/2e5x
in
another.
On
each
of
these
pages,
click
on
the
Ligands
link
from
the
left-hand
sidebar.
As
you
can
see
from
the
above
screenshots,
both
entries
contain
nucleotide-type
ligands.
The
ligand
bound
to
2MJP
is
ANP
(PHOSPHOAMINOPHOSPHONIC
ACID-
ADENYLATE
ESTER).
For
2E5X,
the
bound
nucleotide
is
ITT
(INOSINE
5'-
TRIPHOSPHATE).
Given
that
both
entries
share
a
high
degree
of
structural
similarity
and
also
bind
nucleotide-like
ligands,
do
they
also
share
the
same
or
similar
binding
sites?
It
should
be
possible
answer
that
question
by
looking
at
the
ligand
interaction
results
for
these
two
hetgroups
from
the
corresponding
entries.
Click
on
the
Protein
Data
Bank
in
Europe
http://pdbe.org
PDBeFOLD Tutorial
interactions
link
for
each
of
the
two
ligands
as
shown
above.
This
takes
you
to
the
details
of
the
residue
interactions
in
both
entries
using
another
PDBe
service
(PDBeMotif:
http://pdbe.org/motif/).
2MJP
2E5X
Now
going
back
to
the
residue-by-residue
3D-mapping
results
provided
by
PDBefold,
we
can
make
some
very
interesting
observations.
For
example
ASN
17
and
LYS
20
interacts
with
the
ligand
ANP
in
2MJP,
aligns
with
ASN
9
and
LYS
12
from
PDB
entry
2E5X.
Both
ASN
9
and
LYS
12
are
also
involved
in
interactions
with
ITT
in
a
similar
manner.
In
this
way
if
we
compare
the
rest
of
the
residues
present
in
the
binding
environment
for
ANP
and
ITT,
it
will
be
very
clear
that
both
the
proteins
try
to
adopt
similar
binding
environment
in
order
to
interact
with
a
nucleotide
ligand.
This
observation
also
indicates
that
the
hypothetical
protein
represented
by
the
PDB
entry
2MJP
is
potential
pyrophosphatase
similar
to
the
protein
under
PDB
http://pdbe.org
PDBeFOLD Tutorial
entry
2E5X.
Therefore
in
this
case,
the
function
of
the
protein
molecule
dictates
its
overall
structural
fold
rather
than
its
sequence
identity.
Go
back
to
all
the
results
of
the
structural
comparison
of
2MJP
with
the
PDB.
The
checkbox
shown
on
the
results
page
can
be
used
to
select
entries
to
submit
for
multiple
alignment
from
the
pairwise
alignment
pages.
Check
a
few
of
these
results
and
choose
Submit
for
Multiple
Alignment
from
the
bottom
of
the
results
page.
This
will
do
a
multiple
alignment
between
the
chosen
entries
and
the
results
will
be
shown
similar
to
that
seen
for
the
pairwise
alignment.
You
may
choose
the
View
Superposed
button
to
see
all
your
chosen
entries
aligned.
As
can
be
seen
all
our
chosen
entries
appear
to
share
the
same
overall
structure.
Multiple
Alignments
using
PDBeFOLD
http://pdbe.org
PDBeFOLD Tutorial
From
the
PDBeFOLD
submission
form,
choose
the
radio
button
for
Multiple
Alignment
to
bring
up
the
submission
form
for
multiple
alignments.
The
multiple
alignment
option
provided
by
the
PDBefold
service
is
very
useful
when
you
have
more
than
two
proteins
for
which
you
need
to
compare
structural
folds.
In
query
page
for
multiple
alignment,
type
the
PDB
ID
of
your
choice
and
click
on
the
button
saying
Find
chains.
This
will
upload
the
ID
code
in
the
List
of
entries
column
as
shown
above.
Once
an
entry
is
loaded,
click
on
the
new
entry
button
to
upload
the
next
entry
in
a
similar
way.
The
delete
entry
button
enables
the
user
to
remove
the
highlighted
PDB
ID
code
from
the
list
of
entries.
The
Actualize
button
updates
data
for
the
highlighted
entry
from
the
List
of
entries,
from
the
details
entered
in
the
right
part
of
the
Submission
Form.
Once
all
the
PDB
entries
are
uploaded,
click
on
the
submit
your
query
button.
http://pdbe.org
PDBeFOLD Tutorial
The
multiple
alignment
results
page
shows
structural
alignment
among
the
4
entries
that
were
uploaded
(1b78,
2dvn,
2e5x
and
1v7r).
You
can
also
view
the
superposed
entries
in
a
graphics
viewer
by
pressing
the
Show
Superposed
button.
The
central
domain
region
in
all
proteins
appears
to
be
highly
similar.
Analysis
of
the
residue-by-residue
mapping
data
(similar
way
it
was
done
in
the
pairwise
alignment
before)
also
indicates
high
degree
of
similarity
observed
in
the
active
binding
site
(nucleotide
binding)
of
all
these
proteins.
These
results
obtained
from
PDBefold
service
provide
encouraging
prospects
of
understanding
possible
roles
of
a
hypothetical
protein
or
structural
genomics
proteins
whose
function
is
yet
to
be
determined.
http://pdbe.org
PDBeFOLD Tutorial
OTHER
EXAMPLES
Secondary
structure
alignments
can
often
show
relationships
that
are
not
immediately
obvious
from
sequence
identity
alone.
Here
are
a
few
example
which
you
may
find
interesting.
a) Alpha-lactalbumin
(PDB
entry
1A4V
http://pdbe.org/1a4v).
Start
a
PDBeFOLD
comparison
for
all
entries
in
the
PDB
archive
against
1A4V.
Once
the
results
are
shown,
scroll
to
the
bottom
of
the
page
and
sort
by
%seq
instead
of
Q-score.
Now
scroll
to
the
last
page
and
choose
one
of
the
results
from
the
last
page.
Entry
3cb7
has
31%
sequence
identity
but
70%
structure
similarity
between
the
two
proteins.
Look
at
the
details
of
the
match.
The
structures
are
highly
similar.
http://pdbe.org
PDBeFOLD Tutorial
b) Eosinophil
Major
Basic
Protein
(PDB
entry
1H8U
http://pdbe.org/1h8u).
Do
a
search
for
structural
similarity
for
PDB
entry
1H8U
against
the
whole
PDB
archive.
Once
the
results
are
in,
resort
the
results
by
%seq
identity
as
in
the
previous
example.
Scroll
to
the
last
page.
The
last
hit
in
the
list
is
PDB
entry
1MPU.
This
entry
has
14%
sequence
identity
with
our
query
structure
while
sharing
88%
structural
identity.
Look
at
the
details
of
this
hit
and
view
the
superposed
entries
as
previously.
Scroll
down
the
details
page
and
look
at
the
residue-by-residue
listing.
All
the
CYS
residues
between
the
two
structures
are
conserved
and
at
the
same
equivalent
positions.
These
residues
form
the
disulphide
bonds
in
the
two
structures
that
keep
the
scaffold
of
the
protein
intact.
As
a
matter
of
fact,
both
these
proteins
belong
to
a
large
family
of
sugar-binding
proteins
called
c-type
lectins.
All
c-type
lectins
share
the
same
overall
structure
constrained
by
the
disulphide
bonds,
and
not
all
proteins
in
this
family
actually
bind
sugars.
http://pdbe.org
PDBeFOLD Tutorial
c) PDB
entry
1TIM
and
2O55.
Do
a
pairwise
alignment
between
1TIM
and
2O55
using
the
pairwise
alignment
form.
The
results
will
show
that
the
two
proteins
share
4%
sequence
identity
and
70%
structural
similarity.
Both
proteins
have
minimal
sequence
identity
and
yet
belong
to
the
same
fold
class
(the
TIM
barrel).
This
ends
our
tutorial
on
PDBeFOLD.
We
hope
you
found
this
useful
and
will
be
able
to
use
this
tool
in
your
future
research
and
analysis.
It
is
hopefully
clear
from
the
examples
given
in
this
tutorial
that
fold
space
is
more
restricted
than
sequence
space
and
most
proteins
tend
to
fall
broadly
into
one
of
the
many
fold
classes
already
found
in
the
PDB.
If
you
need
to
get
in
touch
with
the
PDBe
regarding
any
aspect
of
the
programme,
please
email
pdbehelp@ebi.ac.uk
and
we
will
try
to
assist
you
in
any
way
possible.
http://pdbe.org
PDBeFOLD Tutorial