Validation of biomacromolecular structures -
motivation
Radka Svobodová Vařeková
Validation: Why to validate?
Structural biology community found that some published
structures contained serious errors
2
Nightmare before Christmas
Validation: Why to validate?
Garbage in, garbage out
Interesting articles:
• Matthews, B. W. (2007) Five retracted structure reports: inverted or
incorrect? Protein science : a publication of the Protein Society, 16,
1013–6.
• Johnston, C. A., Kimple, A. J., Giguere, P. M., and Siderovski, D. P.
(2008) RETRACTED: Structure of the Parathyroid Hormone Receptor
C Terminus Bound to the G-Protein Dimer Gb1g2. Structure, 16,
1086–1094.
3
Can we still find wrong structures in PDB?
4
Can we still find wrong structures in PDB?
Example: Nipah G attachment glycoprotein (PDB ID 3D12, PNAS)
Contains 30 instances of 11 different carbohydrates, each with one ring
and five chiral atoms.
Results:
• 13 of these ligands have incorrect chirality
• In a few cases, all chiral atoms exhibit incorrect chirality
5
Can we still find wrong structures in PDB?
6
Unfortunately yes
Validation approaches
7
What to validate? How? Software
Geometry (3D) Against tabular values
Proteins and nucl. acids:
WHAT_CHECK, PROCHECK,
PROCHECK-NMR, AQUA,
MolProbity, OOPS
Ligands:
ValLigURL, Mogul, Coot, PHENIX
Topology (2D) Against a template
Proteins and nucl. acids:
-||-
Ligands:
pdb-care, MotiveValidator,
ValidatorDB
1.5 Å
120°
1.7 Å
Compilation:
PDB validation reports
Improved data quality
 Cleaned up mmCIF data
 Standard vocabularies
 Experimental details, binding sites, secondary structure,
antibiotic/inhibitor information, nucleic-acid parameters
 Clean mmCIF files are used in production
Validation reports
 Summary
 Quality vs. all PDB
 Quality vs. entries at
similar resolution
 Overview of residuebased
quality for
every polymer
 Table of ligands that
may need your
attention
Shortcut: pdbe.org/valrep/1cbs
10
Topology validation
(= validation of annotation)
Topology validation – basic terms
11
Residue
 Any component of a biomacromolecule
 Examples: amino acids, nucleotides, saccharides,
ions, …
 In PDB, residue is annotated via “Residue ID” – a
unique 3-letter code
FUC, alpha-L-fucosePHE, phenylalanine
Topology validation – basic terms
12
Types of residues:
Standard residues:
 amino acids
 nucleotides
Non-standard residues:
 modified amino acids and nucleotides
Ligands:
 Chemical compounds which form a complex with a
biomacromolecule (e.g., sugar, drug, heme).
 Also ions are often referred as ligands
Topology validation – basic terms
13
Principles of topology validation:
 Subjects of topology validation are residues
 Validated residue is compared with a model
residue, which has the same Residue ID
 The model residues are taken from a reference
database
 Differences between the model residue and the
validated residue are reported
Topology validation – approach
14
Complication with “shared atoms”:
 When some residues bind together, one of them can lose an
atom:
MAG
FUC
PDB ID 1g1t
H20
• Solution: When we validate a residue, we must include also
its close surrounding
MAG-FUC
Input motif = validated
residue + surrounding
Topology validation - approach
15
Validated motif mapped to the Model Residue
Input PDB entry
Selection of validated
residue and its close
surrounding
Mapping of input motif to the model residue
(via subgraph matching)
Topology validation – types of validation
analyses
16
 Completeness analyses
 Missing atoms
 Missing rings
 Chirality analyses
 Chirality on C atom, metal atom, high bond order
atom, planarity
 Advanced analyses
 Substitution
 Different atom naming
 Foreign atoms
Topology validation – types of validation analyses
Prerequisite: Can we map the validated and the model residue?
Degenerated structure
Correct residue
1IVG_17_7716 (MAN)
Topology validation – types of validation analyses
ERRORS – INCOMPLETE STRUCTURE
18
Missing atom Missing ring
Correct residue
Topology validation – types of validation analyses
ERRORS – CHIRALITY
19
Wrong chirality on
metal atom
1E4M_16_4280 (MAN)
Wrong chirality on C
atoms
MAN AVC
2P7E_0_95 (AVC)
Topology validation – types of validation analyses
ERRORS – CHIRALITY II
20 1BZ0_1_4428 (HEM)
Wrong chirality (planar)
4A2U_2_13909 (CMP)
CMP
Wrong chirality on atom
having high order bonds
HEM
Correct chirality
(Tolerant)
=
Correct residues
+
residues having
only these issues:
• Wrong chirality
(planar)
• Wrong chirality
on atom having
high order
bonds
Topology validation – types of validation analyses
WARNINGS
21
Substitution
Foreign atom
(= atom from neighboring residue)
Different atom
name
O1 instead O5
Correct residue
ADVANCED
WARNINGS:
Alternate
locations
Zero RMSD
with model
Different atom names - example
22
Topology validation
Software tools
PDB care (Lütteke et al., 2006):
• Tool focused on carbohydrates validation
• First application, which implements topology validation
• Performs basic validation analyses (missing atoms, missing rings, wrong chirality)
MotiveValidator (http://ncbr.muni.cz/MotiveValidator) :
• Tool, which allows validation of all residues
• Performs basic validation analyses + reports basic warnings (substitutions, foreign atoms,
different naming)
ValidatorDB (http://ncbr.muni.cz/ValidatorDB):
• Database, containing validation results for all* ligands and non-standard residues in PDB
(weekly updated)
• Performs basic validation analyses + advanced validation analyses (report degenerated
residue, distinguish type of chirality error)
• Reports basic warnings + next warnings (Alternate locations, Zero RMSD with model)
* Except amino acids, nucleotides, and small residues (<7 heavy atoms)
Central European Institute of Technology
Masaryk University
Kamenice 753/5
625 00 Brno, Czech Republic
www.ceitec.muni.cz | info@ceitec.muni.cz
Thank you for your attention