All-Atom Macromolecular Validation: MolProbity is Now Available on Vecura
This update enables structural biologists, crystallographers, and AI researchers to evaluate and validate macromolecular structure quality through a guided workflow inside Vecura, without setting up complex technical infrastructure.
What is MolProbity?
MolProbity is the canonical, all-atom structure-quality validation suite for macromolecular crystallography, cryo-EM, NMR, and AI-predicted protein models. Developed at the Richardson laboratory at Duke University, it evaluates how well a macromolecular model agrees with known stereochemical and geometric standards by probing all-atom contacts, backbone dihedral angles, side-chain rotamers, C-beta geometry, and covalent bonds or angles. It helps users determine the physical plausibility and quality of any 3D molecular structure before final publication or PDB deposition. It is especially useful for structural biologists and AI model developers who need to diagnose local modeling errors or refine predicted protein structures.
What can users do with MolProbity on Vecura?
With MolProbity on Vecura, users can:
- Analyze macromolecular structures in a single call to generate composite quality scores, global percentages (clashscore, Ramachandran outliers, rotamer outliers), and structural tables in one seamless pipeline.
- Diagnose all-atom clashes with physical precision by automatically adding explicit hydrogen atoms to the input PDB file and detecting subtle steric overlaps.
- Identify residue-level geometry outliers by comparing bond lengths, bond angles, and chirality parameters against the classic Engh-Huber and modern conformation-dependent restraint libraries.
- Produce comprehensive per-residue quality tables that compile clash metrics, rotamer classifications, Ramachandran status, and structural outliers into an actionable prioritized list for model rebuilding.
What the output means
The output provides a comprehensive macromolecular quality validation report containing global summary scores, percentage metrics, and residue-level diagnostic tables. It includes the headline MolProbity score, which represents the overall physical plausibility of the model, as well as critical values like the clashscore (counting steric overlaps > 0.4 Å per 1000 atoms), Ramachandran favored and outlier percentages, rotamer outlier rates, C-beta deviations, and bad bonds/angles percentages.
This output should be used to support scientific decision making. It does not replace experimental validation.
Why this matters
The biological function of a macromolecule is intimately tied to its three-dimensional structure. Over the past decade, advancements in structural determination—from high-resolution cryo-EM to deep-learning-based structure prediction engines like AlphaFold—have led to an explosion in newly modeled protein structures. However, these models often suffer from subtle, chemically unrealistic anomalies, such as overlapping electron clouds, strained bonds, or physically impossible side-chain orientations. Identifying these errors is critical, as minor geometric inaccuracies can easily compromise downstream tasks like virtual drug screening, enzyme engineering, or protein-protein interaction modeling.
By bringing MolProbity's gold-standard validation suite onto Vecura, researchers gain access to a highly sensitive and automated debugging tool for macromolecular models. Rather than navigating complex local software compilations, environment paths, or command-line wrappers, scientists can now evaluate structure quality through an intuitive web-based environment. This democratization of structural validation ensures that both experimental and computationally generated models can be rapidly assessed for chemical accuracy, paving the way for safer, more robust biological engineering.
- Developed by: Richardson Laboratory, Duke University
- Source: Official GitHub Repository (rlabduke/MolProbity)
- Reference: Williams et al. (2018) Protein Science and Chen et al. (2010) Acta Crystallographica Section D
Vecura で MolProbity を試す。
モデルワークスペースを開き、ご自身の入力で評価を始めましょう。
