Two parts to successful model
building
• PATTERN
RECOGNITION SKILLS
– Recognizing structural
features in electron density
maps and skeleton maps.
• .a-helix density.
• .b-strand density.
• Recognize a-carbons
positions
• Recognize side chain
density
• BUILDING TOOLS –how
to use Coot
– Initiate trace of protein chain
(“Place helix here”)
– Test sidechain assignments.
• Mutate+autofit
• Select rotamers
– Maintain proper geometry
• “Regularize” bond lengths
and angles, planes, eliminate
steric clash.
• “Real space refine” to adjust
fit to e- density
– Validation tools to detect
disallowed f and y angles.
Basic protein chemistry:
Phi(f) and Psi(Y)
an amino acid residue
Ni+1
C
O
Ca
Og
Cb
N
Oi-1
carbon
nitrogen
oxygen
planar peptides
chiral a-carbons
allowed f,y anglessee Ramachandran plot
Lowest energy f,y angles correspond to
a-helices and b-sheets
b-sheet
a-helix
Ramachandran plot
Lets focus on recognizing helix and strand features in electron density maps.
Be able to recognize a helix from different
perspectives
90o
Viewed down helical axis
Viewed perpendicular to
helical axis
The hole through the center of a helix is the most
distinctive feature of a-helix density.
Viewed down helical axis
Viewed perpendicular to
helical axis
Often it is easier to recognize
helical density when viewed down
the helical axis due to the
distinctive hole through the
center of the helix.
But, be sure to view both perspectives when
modeling an a-helix.
Incorrect
When viewed down
the helical axis. Both
orientations of the
helix appear to fit the
electron density OK.
But, when viewed
perpendicular to the
helical axis, it
becomes clear that
only one direction of
the helix fits the
carbonyl bumps in the
electron density.
Correct
Helix directionality
C-terminus
Backbone
oxygens
10
9
But, electron density isn’t
labeled with residue
numbers.
8
7
6
5
Backbone
nitrogens
4
3
2
1
N-terminus
Protein models are always
numbered from N-terminus
to C-terminus as shown
here. So it is easy to tell the
directionality of a helix.
What structural features are
present in the electron
density map to help you
determine which direction to
place the helix?
Carbonyl oxygens point to C-terminus
C-terminus
10
9
8
Look for
carbonyl
bumps
7
6
5
4
3
2
1
Cb point to the N-terminus like the branches of a Christmas tree.
Cb
Cb
point
to
N-terminus
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb point to the N-terminus like the branches of a Christmas
tree.
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Cb
Which way is the C-terminus
pointing?
Test 2: Which way is the Cterminus pointing ?
b-strands appear as parallel tubes
of density spaced 4.8 Å apart.
b-strands are never alone.
STRAND 1
4.8 Å
STRAND 2
4.8 Å
STRAND 3
4.8 Å
STRAND 4
b-strands also have directionality
From this perspective, side chains of successive residues alternate
in and out of the plane of this page.
N-terminus
1
3
2
5
4
7
6
9
8
C-terminus
10
b-strands viewed from different perspectives
From this perspective, side chains of successive residues alternate
in and out of the plane of this page.
N-terminus
1
3
2
1
5
4
3
7
6
5
90o
9
8
7
C-terminus
10
9
N-terminus
C-terminus
2
4
6
8
10
From this perspective, side chains of successive residues alternate up and down.
Often it is easier to recognize a beta strand by this distinctive zig-zag pattern than by the
pattern shown above.
But, be sure to view both perspectives when
modeling a b-strand.
When viewed using the zig-zag perspective. Both orientations of the strand appear to fit
the electron density OK.
correct
incorrect
But, when viewed perpendicular to the zig-zag perspective, it becomes clear that only
one direction of the strand fits the carbonyl bumps in the electron density.
Proteinase K (Tritirachium album)
amino acid sequence
001_MAAQTNAPWG_LARISSTSPG_TSTYYYDESA_GQGSCVYVID
041_TGIEASHPEF_EGRAQMVKTY_YYSSRDGNGH_GTHCAGTVGS
081_RTYGVAKKTQ_LFGVKVLDDN_GSGQYSTIIA_GMDFVASDKN
121_NRNCPKGVVA_SLSLGGGYSS_SVNSAAARLQ_SSGVMVAVAA
161_GNNNADARNY_SPASEPSVCT_VGASDRYDRR_SSFSNYGSVL
201_DIFGPGTSIL_STWIGGSTRS_ISGTSMATPH_VAGLAAYLMT
241_LGKTTAASAC_RYIADTANKG_DLSNIPFGTV_NLLAYNNYQA
Assigning the sequence
Find a stretch of 5-10
residues with well defined side
chain density.
Find which amino acid best fits
the density by trial and error.
(Mutate & Autofit)
Keep in mind some residues
are isosteric. For example
threonine and valine.
Phe - Thr - Ala- Ser
or
or
Val
Cys
Check the proteinase K amino
acid sequence for a matching
sequence of residues.
Proteinase K (Tritirachium album)
amino acid sequence
001_MAAQTNAPWG_LARISSTSPG_TSTYYYDESA_GQGSCVYVID
041_TGIEASHPEF_EGRAQMVKTY_YYSSRDGNGH_GTHCAGTVGS
081_RTYGVAKKTQ_LFGVKVLDDN_GSGQYSTIIA_GMDFVASDKN
121_NRNCPKGVVA_SLSLGGGYSS_SVNSAAARLQ_SSGVMVAVAA
161_GNNNADARNY_SPASEPSVCT_VGASDRYDRR_SSFSNYGSVL
201_DIFGPGTSIL_STWIGGSTRS_ISGTSMATPH_VAGLAAYLMT
241_LGKTTAASAC_RYIADTANKG_DLSNIPFGTV_NLLAYNNYQA
Some amino acids have distinctive shapes, others are isosteric.
When in doubt, consider the protein environment.
Load Coordinates
• File -> Open Coordinates--browse window
opens
– Click “filter” (will show only .pdb files)
– Click “sort by date” (will place the lastest
coordinates at the top of browser)
– Select the .pdb file (e.g. sawaya-coot-0.pdb)
Open Coordinates
Auto Open MTZ
Open MTZ, mmcif, fcf, or phs
Load Structure Factors (map)
Open Coordinates
Auto Open MTZ
Open MTZ, mmcif, fcf, or phs
• File -> Open mtz, mmcif, fcf, or phs
– Click “filter” (will show only .mtz or .hkl or
.phs files)
– Click “sort by date” (will place the lastest
.phs file at the top of browser)
– Select the .phs file siras-pcmbs.phs
siras-pcmbs.phs
Zoom in on a distinctive side chain.
Calculate ->Model/Fit/Refine
Mutate & Autofit
Select amino acid type
If rotamer is incorrect, choose
another
Rotamers
Cg
Cb
Ca
C
N
O
energetically preferred rotation angles about single bonds in side chains
H
H
H
Cg
H
O=C
N
H
N
O=C
O=C
N
H
Newman Projections
A rotamer is energetically favorable because it is one of 3 possible staggered conformations.
Cg
H
HH
O=C
Nearly eclipsed. Unfavored.
H N
Sample different rotamers
Accept
Real Space Refinement can tidy up.
If you don’t like it, Reject it.
If you accidentally accepted, then, “undo”.
If molecule explodes, adjust
refinement weight.
5.0
Lower numbers tighten geometric restraints
Move to next residue. Mutate &
Auto Fit.
What is the next residue?
What is the sequence of these
three amino acids?
FTA
Or
FVA
?????
Proteinase K (Tritirachium album)
amino acid sequence
001_MAAQTNAPWG_LARISSTSPG_TSTYYYDESA_GQGSCVYVID
041_TGIEASHPEF_EGRAQMVKTY_YYSSRDGNGH_GTHCAGTVGS
081_RTYGVAKKTQ_LFGVKVLDDN_GSGQYSTIIA_GMDFVASDKN
121_NRNCPKGVVA_SLSLGGGYSS_SVNSAAARLQ_SSGVMVAVAA
161_GNNNADARNY_SPASEPSVCT_VGASDRYDRR_SSFSNYGSVL
201_DIFGPGTSIL_STWIGGSTRS_ISGTSMATPH_VAGLAAYLMT
241_LGKTTAASAC_RYIADTANKG_DLSNIPFGTV_NLLAYNNYQA
Extend N and C termini
one amino acid at a time
• Center on
the N or Cterminus of
the helix
• Click on
“Add
Terminal
Residue”
• Accept or
drag to
better
location.
Remember
• Do not use maximize button to expand the Coot window
to full screen mode. It will hide “pop up” dialog boxes.
– Coot will be waiting for a response, but you’ll never see the
question because it is hidden behind the full screen window.
– Instead, stretch window by dragging corner.
X
NO!
X
X
drag
corner
Save coordinates frequently
or suffer set backs.
Saving first set of coordinates.
• File menu
– Save coordinates
• Select which coordinate
set you want to save.
– 1 Helix
• Auto suggestion:
– Helix-coot-0.pdb
• Change to:
– sawaya-coot-0.pdb
version number
1
Updated coordinates? Save with
incremented version number.
• Select save.
– 1 sawaya-coot-0.pdb
1 sawaya-coot-0.pdb
• Auto suggestion:
sawaya-coot-1.pdb
• Accept .
• Next time:
– sawaya-coot-2.pdb,
– sawaya-coot-3.pdb, etc.
sawaya-coot-1.pdb
Save coordinates frequently
1) Every 5 minutes.
2) When you have done some modeling that
you are especially pleased with.
3) When you are fear that the next step is
going to destroy your previous work.
Results
•
•
•
•
•
•
•
•
•
sawaya1-coot-0.pdb
sawaya1-coot-1.pdb
sawaya1-coot-2.pdb
sawaya1-coot-3.pdb
sawaya1-coot-4.pdb
sawaya1-coot-5.pdb
sawaya1-coot-6.pdb
sawaya1-coot-7.pdb
sawaya1-coot-8.pdb
The last version number of
each baton build session
represent your best effort
at modeling that segment
of the protein chain.
Next week, we will
concatenate these
segments of chain into one
file and refine them,
calculate Rwork and Rfree.
To control objects
Rotate
around x or y
Translate
in x or y