WORK IN PROGRESS

Content:

• Introduction

• Looking at initial maps
• Deposited processed data
• (Re-)processed raw data (traditional, isotropic)
• (Re-)processed raw data (anisotropic, STARANISO)
• (Re-)processed raw data (traditional, isotropic) - F(early)-F(late) radiation-damage maps

• Initial (re-)refinement

• Footnotes

---

Introduction

• RCSB PDBe PDBj
• Pallesen, J., Wang, N., Corbett, K.S., Wrapp, D., Kirchdoerfer, R.N., Turner, H.L., Cottrell, C.A., Becker, M.M., Wang, L., Shi, W. and Kong, W.P., 2017. Immunogenicity and structures of a rationally designed prefusion MERS-CoV spike antigen. Proceedings of the National Academy of Sciences, 114(35), pp.E7348-E7357.

What was the reason to pick this structure? It was one of the only 2 PDB entries with diffraction data deposited (thanks to the authors!) - so we might as well pick this one to also show how to (re-)process the raw diffraction images.

We can fetch the original, deposited data using

  fetch_PDB 5VZR

which will give use files

5VZR/5vzr.pdb
5VZR/5vzr.mtz

and various other diagnostics. All subsequent commands are run within the 5VZR subdirectory created by this command.

Note: all maps are shown at rms=1.0 (2mFo-DFc) and rms=3.5 (mFo-Dfc) levels unless stated otherwise.

---

Looking at initial maps

There are multiple ways of looking at electron density maps for a given PDB entry. The easiest is to use the tools from one of the wwPDB sites, e.g. https://www.rcsb.org/3d-view/5VZR?preset=electronDensityMaps. But if you are more familiar with using Coot, you should be able to load both model and electron density (2mFo-DFc) and difference (mFo-Dfc) maps from within the Coot interface. You could also go directly to PDB-REDO and look at the available models and maps.

Often, PDB entries will contain the map-coefficients of the maps as originally seen by the depositor/author (items _refln.pdbx_FWT, _refln.pdbx_PHWT, _refln.pdbx_DELFWT and _refln.pdbx_DELPHWT in the structure factor mmCIF file). ¹

Here we are going to use the so-called "MapOnly" mode in BUSTER, namely running

  refine -M MapOnly -p 5vzr.pdb -m 5vzr.mtz -d MapOnly | tee MapOnly.lis
  cd MapOnly
  coot --pdb refine.pdb --auto refine.mtz

So we are not doing any refinement of the deposited model (apart from the overall scaling model) and compute the different types of maps.

Deposited processed data

What can we see (in order of largest difference map peaks)? Some examples showing a few possible, small corrections:

Location	Maps	Remarks
A223		some density for additional residues (A224-A235 not yet modelled)
H61, H65		maybe alternate loop conformation and signs of radiation damage
H206		broken disulfide bond due to radiation damage
L188		different rotamer with additional water?
H80		different/alternate rotamers
B11		additional water
L194		broken disulfide bond due to radiation damage
H43		glycerol (GOL) molecule?
L103		alternate conformation

(Re-)processed raw data (traditional, isotropic)

Location	Maps	Remarks
A223		some density for additional residues (A224-A235 not yet modelled)
H61, H65		maybe alternate loop conformation and signs of radiation damage
H206		broken disulfide bond due to radiation damage
L188		different rotamer with additional water?
H80		different/alternate rotamers
B11		additional water
L194		broken disulfide bond due to radiation damage
H43		glycerol (GOL) molecule?
L103		alternate conformation

(Re-)processed raw data (anisotropic, STARANISO)

Location	Maps	Remarks
A223		some density for additional residues (A224-A235 not yet modelled)
H61, H65		maybe alternate loop conformation and signs of radiation damage
H206		broken disulfide bond due to radiation damage
L188		different rotamer with additional water?
H80		different/alternate rotamers
B11		additional water
L194		broken disulfide bond due to radiation damage
H43		glycerol (GOL) molecule?
L103		alternate conformation

(Re-)processed raw data (traditional, isotropic) - F(early)-F(late) radiation-damage maps

When there is enough multiplicity/redundancy in the data, autoPROC can merge distinct parts of the dataset into so-called "early" and "late" datasets. When such a reflection file (truncate-unique.mtz from autoPROC) is given to BUSTER it will create F(early)-F(late) difference Fourier maps that should show positive peaks where there was something at the beginning of data collection (atoms, electrons) and negative peaks where something appeared towards the end of data collection. So they are easy visualisations of potential radiation damage effects.

We also get a list of the highest positive peaks close to existing atoms, which would be those atoms/residues whihc are most likely damaged. Looking at those peaks (above 4.5 rms) here

Peak         Closest atom in aP_iso2.MapOnly/refine.pdb
[rms]                                             Distance (<= 1.0 )
------------------------------------------------------------------------
 10.10  <=>   SG  CYS B 194  (  1.00 31.27)  :       0.27
  7.78  <=>   SG  CYS L  88  (  1.00 18.68)  :       0.18
  7.65  <=>   SD  MET B  33  (  1.00 26.31)  :       0.36
  7.56  <=>   SG  CYS B  88  (  1.00 19.91)  :       0.48
  7.53  <=>   SG  CYS B  23  (  1.00 23.85)  :       0.38
  7.26  <=>   SG  CYS L  23  (  1.00 24.13)  :       0.32
  7.04  <=>   SD  MET L  33  (  1.00 26.26)  :       0.40
  6.76  <=>   SG  CYS B 134  (  1.00 21.97)  :       0.51
  6.49  <=>   SG  CYS L 194  (  1.00 40.76)  :       0.25
  6.05  <=>   SG  CYS H  22  (  1.00 17.90)  :       0.50
  5.93  <=>   SG  CYS A  22  (  1.00 20.65)  :       0.49
  5.77  <=>   SG  CYS A 206  (  1.00 24.34)  :       0.24
  5.68  <=>   SG  CYS A 140  (  1.00 22.55)  :       0.46
  5.64  <=>   SG  CYS H  92  (  1.00 15.67)  :       0.38
  5.32  <=>   SG  CYS A  92  (  1.00 17.29)  :       0.57
  5.24  <=>   SG  CYS H 140  (  1.00 22.66)  :       0.44
  5.21  <=>   O   HOH H 312  (  1.00 20.89)  :       0.56
  5.20  <=>   O   HOH H 365  (  1.00 23.87)  :       0.49
  5.19  <=>   CD  GLU L 123  (  1.00 47.18)  :       0.51
  5.17  <=>   O   ASP L 165  (  1.00 21.03)  :       0.70
  5.10  <=>   O   HOH H 426  (  1.00 30.93)  :       0.57
  4.98  <=>   OD2 ASP H  72  (  1.00 26.72)  :       0.65
  4.97  <=>   OD1 ASN H 207  (  1.00 20.02)  :       0.76
  4.93  <=>   OD1 ASP A 100C (  1.00 22.25)  :       0.68
  4.92  <=>   OE2 GLU L  27  (  1.00 50.86)  :       0.52
  4.85  <=>   OD2 ASP A 218  (  1.00 34.32)  :       0.49
  4.76  <=>   CG  ASP B  70  (  1.00 21.87)  :       0.76
  4.70  <=>   O   HOH B 393  (  1.00 19.92)  :       0.47
  4.69  <=>   O   HOH H 344  (  1.00 28.64)  :       0.62
  4.69  <=>   O   VAL H 217  (  1.00 15.41)  :       0.65
  4.69  <=>   CG2 THR H 205  (  1.00 15.67)  :       0.58
  4.68  <=>   CE BMET A  80  (  0.55 18.23)  :       0.52
  4.68  <=>   OG1 THR H 154  (  1.00 20.07)  :       0.66
  4.67  <=>   OD1 ASN L 137  (  1.00 18.09)  :       0.20
  4.67  <=>   SD  MET H 100E (  1.00 22.30)  :       0.61
  4.63  <=>   O   HOH A 453  (  1.00 40.93)  :       0.91
  4.62  <=>   OG  SER A 120  (  1.00 21.87)  :       0.84
  4.62  <=>   OD1 ASP L   1  (  1.00 33.52)  :       0.48
  4.62  <=>   OE1 GLU H 148  (  1.00 26.11)  :       0.72
  4.58  <=>   SG  CYS L 134  (  1.00 26.62)  :       0.77
  4.58  <=>   CG  ASP L 170  (  1.00 29.17)  :       0.61
  4.56  <=>   OD2 ASP L  27C (  1.00 30.67)  :       0.71
  4.53  <=>   OD1 ASP A 181  (  1.00 39.08)  :       0.32
  4.52  <=>   OD1 ASP A 218  (  1.00 28.37)  :       0.64

shows all the usual suspects: sulfurs CYS/MET residues mainly, but also some close to GLU/ASP carboxylates.

Looking at those maps in Coot (take columns F_early-late and PHI_early-late in the refine.mtz output file to create a difference map):

B194
B88
L88
B23
L23
A10

---

Initial (re-)refinement

We are starting by using the anisotropic (STARANISO) analysed data from autoPROC, because:

• the scaling/merging statistics are consistently better than those for the original processing (deposited data)
• in our experience the data from STARANISO is neutral or (in good cases) beneficial during refinement, leading to clearer maps (your mileage might vary)
• for comparisons we will later) also look at refinements against the original (deposited) data as well as against data without any or with a (traditional) isotropic resolution cut

---

Footnotes

• *1: this would allow you to look at the maps exactly as the authors saw them and used for building the deposited model