# What Are Data Standards in Structural Biology? A Beginner's Guide to PDBx/mmCIF, EMDB, and Validation Metrics

Structural biology depends on shared formats and quality metrics that let researchers deposit, retrieve, and compare three-dimensional models of biological molecules. Without these standards, a structure solved in one laboratory could not be interpreted reliably by another group, and the growing archives of macromolecular data would become unusable. This article explains the PDBx/mmCIF format used by the Protein Data Bank, the map formats used by the Electron Microscopy Data Bank, and the validation metrics that indicate whether a deposited structure is trustworthy. The practical goal is to give students and researchers the working knowledge needed to open a structure file, assess its quality, and decide whether it is suitable for downstream analysis such as docking studies or structure-based drug design.

## The Role of Data Standards in Structural Biology

Data standards in structural biology are the agreed-upon rules for describing molecular structures in a way that computers can parse and humans can interpret. They define how atomic coordinates, experimental details, and quality indicators are recorded in a file. The Protein Data Bank (PDB) began as a small repository for crystallographers and has grown into a worldwide resource that serves the entire structural biology community. This growth depended on the community developing standards and policies for data deposition and exchange. The history of the PDB shows that standards are living agreements that evolve as new experimental methods emerge and as the complexity of deposited structures increases.

The practical consequence of these standards is interoperability. When a researcher downloads a structure file, the format must carry enough information for software to display the molecule, for validation tools to assess its quality, and for other scientists to reproduce the analysis. The worldwide PDB network now includes collaborating biophysics communities that have created interoperating data resources. This network expands the capabilities of structural biology and enables the determination and archiving of increasingly complex structures, including large molecular machines and integrative models that combine data from multiple experimental techniques.

For a beginner, the most important point is that a structure file is also a list of atoms. It is a structured record that includes the experimental method, the resolution or map quality, the refinement statistics, and the validation metrics. Learning to read these records is a core skill in structural bioinformatics.

## PDBx/mmCIF: The Core Format for Macromolecular Structures

The PDBx/mmCIF format is the current standard for depositing and distributing macromolecular structures in the Protein Data Bank. The name combines two ideas: PDBx refers to the extension of the original PDB format, and mmCIF stands for macromolecular Crystallographic Information Framework. The format is based on a dictionary of defined data items, each with a specific name and meaning. This dictionary approach makes the format extensible, so new experimental methods and new types of data can be added without breaking existing files.

### Why the Original PDB Format Became Insufficient

The original PDB format used fixed-width columns to record atomic coordinates. Each line in the file had a specific length, and each field occupied a defined set of columns. This format worked well for small proteins and nucleic acids, but it had serious limitations. The fixed-width layout could not easily accommodate structures with more than 99,999 atoms, and it lacked a formal way to describe the relationships between different parts of a structure. As structural biology moved toward larger complexes, such as ribosomes and viral capsids, the format became a bottleneck.

The transition to PDBx/mmCIF solved these problems by using a tag-value syntax. Each data item is written as a tag followed by its value, and related items are grouped into categories. This design allows files to grow without arbitrary size limits and makes it possible for software to extract specific data items without parsing the entire file. The format also supports the detailed annotation needed for modern validation reports.

### How to Read a PDBx/mmCIF File

A PDBx/mmCIF file is organized into categories that begin with an underscore. For example, the category `_atom_site` contains the atomic coordinates, and the category `_refine` contains the refinement statistics. Each category has a list of attributes, and the data are presented in a table-like layout. The first line of a category block lists the attribute names, and subsequent lines list the values for each atom or entity.

When you open a PDBx/mmCIF file in a text editor, you will see lines that look like this:

```
_atom_site.group_PDB
_atom_site.id
_atom_site.type_symbol
_atom_site.label_atom_id
_atom_site.label_comp_id
_atom_site.label_asym_id
_atom_site.Cartn_x
_atom_site.Cartn_y
_atom_site.Cartn_z
```

The values below these headers give the element symbol, the atom name, the residue name, the chain identifier, and the Cartesian coordinates in angstroms. The file also contains metadata about the experiment, including the resolution, the R-factor, and the software used for refinement.

For practical work, most users do not read these files directly. Instead, they use software such as PyMOL, ChimeraX, or the web-based tools provided by the PDB. However, understanding the file structure helps when you need to troubleshoot a parsing error or when you want to extract specific data items using a scripting language such as Python.

### The Transition Timeline and Current Status

The worldwide PDB officially transitioned to PDBx/mmCIF as the primary deposition format. New depositions are required to use this format, and the legacy PDB format is no longer accepted for new entries. The archive continues to provide legacy-format files for older entries, but these are generated from the mmCIF master files. For any new structure, you should expect to work with PDBx/mmCIF.

The practical implication for researchers is that older software tools may not parse PDBx/mmCIF files correctly. Before using a tool for structure analysis, check its documentation to confirm that it supports the current format. Many actively maintained tools have added mmCIF support, but some legacy scripts still assume the fixed-width PDB format.

## EMDB: Map Formats for Electron Microscopy

The Electron Microscopy Data Bank (EMDB) is the archive for three-dimensional maps produced by electron microscopy, particularly cryo-electron microscopy (cryo-EM). Unlike crystallography, which produces atomic coordinates directly from diffraction data, cryo-EM produces a three-dimensional density map that must be interpreted to build an atomic model. The map is a grid of density values, and the format must preserve the grid dimensions, the voxel spacing, and the density values in a way that visualization software can render.

### Map File Formats and Their Characteristics

The most common format for EMDB maps is the MRC format, named after the Medical Research Council in the United Kingdom where it was developed. The MRC format stores the density map as a three-dimensional array of floating-point numbers, along with a header that records the grid dimensions, the voxel size, and the origin of the map. The format is simple and widely supported by cryo-EM software.

Another format you may encounter is CCP4, which is used primarily in crystallography but is also compatible with many EM tools. The CCP4 format is similar to MRC in its basic layout, and many software packages can read both. The choice of format is often determined by the software used for refinement and visualization, so you should check the requirements of your analysis pipeline.

### Map Resolution and Interpretation Limits

The resolution of a cryo-EM map determines what features can be seen. At resolutions around 4 angstroms, the secondary structure elements such as alpha helices and beta sheets are visible, but side chains are not clearly resolved. At resolutions around 3 angstroms, side chains become visible, and at resolutions better than 2.5 angstroms, the map approaches the quality of a crystallographic structure. The resolution is reported in the EMDB entry and should be the first thing you check when deciding whether a map is suitable for your analysis.

The interpretation of a map is not automatic. Building an atomic model into a density map requires specialized software and careful judgment. The model must fit the density well, and the validation metrics must indicate that the fit is reasonable. A map with a reported resolution of 4 angstroms may be useful for understanding the overall architecture of a complex, but it is not suitable for studying detailed interactions between side chains.

### How EMDB Entries Are Organized

Each EMDB entry has an accession code that begins with EMD, followed by a number. The entry page provides the map file for download, along with metadata about the specimen, the microscope, the data collection parameters, and the reconstruction software. The entry also includes a validation report that summarizes the map quality and, if an atomic model has been deposited, the fit of the model to the map.

When you download an EMDB map, you should also check whether a corresponding atomic model has been deposited in the PDB. Many cryo-EM structures have both a map in EMDB and a model in PDB, and the two entries are cross-referenced. The model provides the atomic interpretation, while the map provides the experimental evidence. Both are needed for a complete understanding of the structure.

## Validation Metrics: How to Assess Structure Quality

Validation metrics are numerical indicators that tell you how reliable a structure is. They are calculated by software that checks the geometry of the model, the fit of the model to the experimental data, and the agreement between different parts of the structure. The two main categories are crystallographic metrics, which apply to X-ray structures, and geometric metrics, which apply to all structures regardless of the experimental method.

### R-free and R-work for Crystallographic Structures

The R-factor is a measure of how well the atomic model explains the observed diffraction data. It is calculated by comparing the measured intensities of the diffraction spots with the intensities calculated from the model. A lower R-factor indicates a better fit. However, the R-factor can be artificially improved by overfitting, which means adding too many parameters to the model to force a better fit to the data.

To guard against overfitting, crystallographers use the free R-factor, commonly called R-free. The calculation of R-free involves setting aside a small percentage of the diffraction data before refinement begins. This test set is never used during refinement, so it provides an independent check of the model. The R-free is calculated by comparing the model with the test-set data. A large gap between R-work and R-free suggests that the model is overfit and may not be reliable.

For a beginner, the practical rule is to check both R-work and R-free. A structure with an R-free below 0.25 is generally considered good, while values above 0.30 raise concerns. The gap between R-work and R-free should be small, typically less than 0.05. These thresholds are not absolute, and the expected values depend on the resolution of the data, but they provide a useful starting point for assessment.

### MolProbity Scores and Geometric Validation

MolProbity is a validation tool that checks the geometry of a structure by looking for problems such as steric clashes, unusual bond lengths, and unfavorable backbone conformations. The output includes a clash score, a Ramachandran plot analysis, and a rotamer analysis. The clash score counts the number of severe overlaps between atoms that should not be in contact. The Ramachandran analysis shows whether the backbone dihedral angles of each residue fall in the favored or allowed regions of conformational space.

The MolProbity score combines several of these geometric checks into a single number that can be compared across structures. A lower MolProbity score indicates better geometry. The score is reported in the PDB validation report, and it is also used by the PDB to flag structures that may have problems. When you are comparing multiple structures of the same protein, the MolProbity score can help you choose the one with the best geometry.

### Validation Reports and the wwPDB OneDep System

The worldwide PDB (wwPDB) provides a validation report for every deposited structure. The report is generated automatically when the structure is deposited, and it includes the R-free, the MolProbity score, the clash score, and other metrics. The report also includes a summary of the experimental data and a list of any issues that were flagged during validation.

The validation report is available on the PDB entry page, and you should always download and review it before using a structure in your research. The report gives you a quick overview of the structure quality, and it highlights any areas of concern. If the report flags serious problems, you should consider whether the structure is suitable for your purposes or whether you should look for an alternative.

## At a Glance: Key Formats and Metrics

The table below summarizes the main formats and metrics you will encounter when working with structural data. Use this table as a quick reference when you are deciding which files to download and which metrics to check.

| Data Type | Format or Metric | What It Records | How to Assess It |
| --- | --- | --- | --- |
| Atomic coordinates | PDBx/mmCIF | Atom positions, connectivity, experimental metadata | Check the resolution and R-free in the `_refine` category |
| Legacy coordinates | PDB format | Atom positions in fixed-width columns | Use only for older entries, prefer mmCIF for new work |
| Cryo-EM density map | MRC or CCP4 | Three-dimensional density grid | Check the reported resolution and the map-to-model fit |
| Crystallographic fit | R-work and R-free | Agreement between model and diffraction data | R-free below 0.25 is good, large R-work/R-free gap indicates overfitting |
| Geometric quality | MolProbity score | Clashes, bond geometry, backbone conformations | Lower score is better, compare with other structures of similar resolution |
| Overall validation | wwPDB validation report | Summary of all quality indicators | Review the report before using the structure |

## Practical Workflow for Using Structural Data

The workflow below describes the steps you should follow when you need to use a structure from the PDB or EMDB. The goal is to ensure that you start with a reliable structure and that you understand the limitations of the data before you proceed with downstream analysis.

### Step 1: Identify the Structure You Need

Start by searching the PDB for structures of your molecule of interest. The search interface allows you to filter by organism, experimental method, resolution, and other criteria. If you are studying a protein that has been solved multiple times, you will need to decide which structure to use. The search results page shows the resolution and the release date, which are useful for an initial screening.

For cryo-EM structures, search the EMDB as well as the PDB. Some structures have a map in EMDB but no model in PDB, and some have both. The EMDB entry page will tell you whether a model is available and will link to the corresponding PDB entry.

### Step 2: Download the Files and Review the Validation Report

Download the PDBx/mmCIF file for the structure and the validation report. Open the validation report and check the key metrics: the resolution, the R-free (for X-ray structures), the MolProbity score, and the clash score. If the structure is from cryo-EM, check the reported map resolution and the model-to-map fit.

If the validation report flags serious problems, consider whether the structure is still useful for your purposes. For some analyses, such as comparing the overall fold of a protein, a structure with moderate quality may be acceptable. For other analyses, such as docking a small molecule into a binding site, you need a structure with good geometry and clear density for the binding site.

### Step 3: Load the Structure into Visualization Software

Use a molecular visualization program such as PyMOL or ChimeraX to load the PDBx/mmCIF file. These programs handle the format automatically, so you do not need to parse the file manually. Once the structure is loaded, you can inspect the molecule, measure distances, and generate images.

For cryo-EM structures, load both the map and the model. The map is displayed as a surface or mesh, and the model is displayed as atomic sticks or ribbons. The fit of the model to the map can be assessed visually, and most visualization programs provide tools for adjusting the map contour level.

### Step 4: Check the Completeness of the Model

Inspect the model for missing residues or atoms. Many structures have disordered regions that could not be resolved in the experiment, and these regions are omitted from the model. The PDBx/mmCIF file records which residues are present, and the validation report may note missing regions. If your analysis depends on a region that is missing from the model, you will need to find another structure or use modeling to build the missing parts.

### Step 5: Document the Structure and Its Quality

Record the accession code, the resolution, the validation metrics, and the date you downloaded the structure. This documentation is important for reproducibility. If you publish results based on the structure, you must cite the original paper and the PDB entry. The PDB entry page provides the citation information.

## Options and Tradeoffs in Structure Selection

When multiple structures of the same molecule are available, you must decide which one to use. The choice depends on your research question and the quality of the available structures.

### Resolution versus Completeness

A structure with higher resolution is generally better, but it may be less complete. For example, a high-resolution crystal structure may lack a flexible loop that is important for your analysis, while a lower-resolution cryo-EM structure may include the full-length protein. You must weigh the benefits of better atomic detail against the benefits of a more complete model.

### Experimental Method Considerations

X-ray crystallography and cryo-EM produce different types of data. Crystal structures are solved at resolutions that can be very high, but the protein is studied in a crystal lattice, which may not reflect the solution state. Cryo-EM structures are determined in a more native-like environment, but the resolution is often lower. For some questions, such as how a protein binds to a ligand, a high-resolution crystal structure is preferable. For other questions, such as how a protein complex assembles in solution, a cryo-EM structure may be more relevant.

### The Role of Molecular Dynamics Simulations

Molecular dynamics (MD) simulations can complement experimental structures by providing information about conformational changes and dynamics. However, the lack of available data sets and structures in most published computational work has limited other researchers' use of these models. Recent initiatives favor the deposition of ensembles and structures to accompany publications, which improves data reuse. The lack of uniform metadata collection, formats, and deposited data limits the impact of MD results and their use by communities that are not experts in MD. When you use MD simulations in your own work, you should follow the emerging standards for data deposition and metadata to make your results usable by others.

## Records and Measurements for Reproducible Analysis

Reproducibility in structural bioinformatics depends on keeping accurate records of the data you use and the steps you take in your analysis. The following practices will help you maintain the traceability of your work.

### Record the Accession Codes and Download Dates

For every structure you use, record the PDB or EMDB accession code and the date you downloaded the file. Structural databases are updated regularly, and a file downloaded today may differ from the same entry downloaded a year ago. The download date is essential for reproducing your analysis at a later time.

### Record the Software Versions

The software you use for visualization, validation, and analysis changes over time. Record the version numbers of all programs in your analysis pipeline. If you use a script to process structures, save the script and note the version of the scripting language and the libraries you used.

### Record the Validation Metrics

Save the validation report for each structure you use. The report provides a snapshot of the structure quality at the time of deposition. If the structure is updated, the validation report will change, so keeping a copy of the original report is important for documenting your analysis.

### Use Workflow Tools for Complex Analyses

For complex analyses that involve multiple steps, consider using a workflow management system. The [Galaxy Training Network](https://training.galaxyproject.org/) provides accessible workflow training and analysis tutorials that can help you build reproducible pipelines. The [nf-core documentation](https://nf-co.re/docs) describes community standards for pipeline usage and configuration, which can help you run reproducible analyses at scale. These tools are particularly useful when you need to process many structures or when you want to share your analysis with collaborators.

## Common Failure Patterns in Structure Interpretation

Beginners often make predictable mistakes when working with structural data. Recognizing these patterns can help you avoid them.

### Ignoring the Validation Report

The most common mistake is to download a structure and use it without checking the validation report. A structure with poor geometry or a high R-free may still be useful for some purposes, but you need to know about the problems before you draw conclusions. Always review the validation report before using a structure.

### Confusing Map Resolution with Model Quality

For cryo-EM structures, the map resolution is not the same as the model quality. A map with a good resolution may have a model that fits poorly, and a map with a moderate resolution may have a well-built model. The validation report includes metrics for the model-to-map fit, and you should check these before using the model.

### Using a Structure Outside Its Applicable Range

Every structure is determined under specific conditions, and the model represents the molecule in that state. A structure of a protein in its unbound state may not be suitable for docking studies that assume a particular conformation. Check the experimental conditions and the biological context before using a structure for a new purpose.

### Overinterpreting Low-Resolution Features

At low resolution, the positions of side chains are not reliable, and even the backbone may be uncertain in some regions. Avoid making claims about specific interactions based on a low-resolution structure. If your analysis depends on detailed atomic interactions, use a structure with a resolution better than 3 angstroms.

### Failing to Account for Missing Regions

Many structures have missing residues, and these regions are not modeled. If you do not check for missing regions, you may assume that a loop or a terminal region is absent when it is simply not resolved. The PDBx/mmCIF file records the residues that are present, and the validation report may note the missing regions.

## Limitations of Structural Data and Interpretation

Structural data have inherent limitations that affect how they can be interpreted. Understanding these limitations is essential for using the data responsibly.

### The Structure Is a Snapshot

A crystal structure or a cryo-EM structure represents one state of the molecule, captured under specific conditions. Proteins are dynamic, and the structure may not reflect the full range of conformations that the molecule adopts in solution. Molecular dynamics simulations can provide information about dynamics, but the simulations themselves depend on the experimental structure as a starting point.

### The Model Is an Interpretation

The atomic model is an interpretation of the experimental data, not a direct observation. Different software packages may produce slightly different models from the same data, and the refinement process involves many choices that affect the final result. The validation metrics provide some indication of the reliability of the model, but they cannot capture all sources of uncertainty.

### The Data May Be Incomplete

The experimental data may not cover all parts of the molecule. Disordered regions, flexible loops, and terminal segments are often missing from the model. The data may also be incomplete in other ways, such as missing ligand molecules or ions that are present in the biological state but not visible in the experiment.

### The Format May Limit the Analysis

The format of the data can limit the types of analysis that are possible. For example, the PDBx/mmCIF format does not directly store information about the dynamics of the molecule, and the MRC format for EM maps does not store information about the experimental conditions. For some analyses, you may need to combine data from multiple sources, and the lack of uniform metadata can make this difficult.

## Quality Controls and Professional Escalation Criteria

Knowing when to seek help or escalate a problem is an important skill in structural bioinformatics. The following criteria indicate situations where you should consult a more experienced colleague or a professional service.

### When the Validation Report Flags Serious Problems

If the validation report indicates serious problems, such as a very high R-free, a large number of clashes, or a poor model-to-map fit, you should not use the structure without understanding the issues. Consult a structural biologist who can help you interpret the report and decide whether the structure is usable.

### When You Cannot Parse the File

If your software cannot parse a PDBx/mmCIF file, the problem may be with the file or with the software. Check the file for errors, and verify that your software supports the current format. If the problem persists, consult the software documentation or seek help from the software support forum.

### When the Structure Does Not Match the Literature

If the structure you downloaded does not match the description in the original paper, there may be a discrepancy between the deposited data and the published results. Check the accession codes and the validation report, and contact the authors if necessary.

### When You Need to Deposit Your Own Data

If you are depositing your own structure, you must follow the deposition guidelines provided by the wwPDB. The deposition process includes validation checks, and you may need to revise your model to address the issues that are flagged. If you are unfamiliar with the deposition process, seek guidance from a colleague who has deposited structures before.

## Safety and Regulatory Context

Structural biology data are used in many applications, including drug development and biotechnology. The standards and validation metrics described in this article are part of a broader framework that ensures the reliability of structural data used in these applications.

### Data Integrity and Reproducibility

The standards for data deposition and validation are designed to ensure that structural data are reliable and reproducible. The wwPDB validation reports provide a transparent record of the quality of each structure, and the deposition process requires that the experimental data be archived along with the model. This transparency is essential for the integrity of the scientific record.

### The Role of Standards in Regulatory Decisions

Structural data are used in regulatory decisions, such as the approval of new drugs. The quality of the structural data can affect the outcome of these decisions, so the validation metrics are important for assessing the reliability of the evidence. Researchers who provide structural data for regulatory purposes must ensure that the data meet the required standards.

### Ethical Use of Structural Data

Structural data are a public resource, and their use should follow ethical guidelines. When you use a structure from the PDB or EMDB, you should cite the original authors and the database entry. You should also respect any restrictions on the use of the data, such as those that apply to structures that are not yet released.

## A Decision Framework for Choosing Between Competing Structures of the Same Target

When multiple structures of the same molecule exist in the PDB or EMDB, the choice of which entry to use can change the outcome of your downstream analysis. A docking study, a structure-based sequence alignment, or a molecular dynamics starting point will inherit the strengths and weaknesses of the chosen structure. The following decision framework gives you a repeatable method for comparing competing entries and documenting your selection rationale.

### Step 1: Define the Intended Use Before Comparing Entries

The first decision is not about the structures themselves but about your analysis goal. Write down what you need from the structure before you open any entry. For example, if you are studying a protein-ligand interaction, you need a structure with the ligand bound and clear density for the binding site. If you are studying conformational changes, you need structures that capture different states of the molecule. If you are building a homology model, you need a template with good geometry and high sequence identity to your target.

This step prevents the common failure of choosing the highest-resolution structure without considering whether it actually contains the features you need. A 1.8 angstrom structure of the apo form will not help you understand ligand binding if the binding site is disordered in the absence of the ligand.

### Step 2: Build a Comparison Table with the Key Metrics

Create a table with one row per candidate structure and columns for the accession code, experimental method, resolution, R-free (for X-ray), MolProbity score, clash score, completeness of the model, and presence of the features you need. The wwPDB validation report provides most of these values directly. For cryo-EM entries, add the map resolution and the model-to-map fit metrics from the EMDB validation report.

The comparison table forces you to look at all candidates side by side instead of fixating on a single metric. A structure with slightly worse resolution but a complete model may be more useful than a higher-resolution structure with large missing regions.

### Step 3: Apply the Resolution and Completeness Tradeoff

For X-ray structures, check the R-free and the gap between R-work and R-free. A large gap suggests overfitting and reduces confidence in the model even if the resolution is good. For cryo-EM structures, check the map resolution and the model-to-map fit. A map with a reported resolution of 3.5 angstroms may still have poorly defined regions that limit the reliability of the model in those areas.

Completeness matters as much as resolution. Compare the residue ranges present in each model. If one structure includes a functionally important loop that another structure lacks, that completeness may outweigh a small resolution disadvantage. The PDBx/mmCIF file records the residue ranges, and the validation report may note missing regions.

### Step 4: Check the Experimental Context

The experimental conditions affect the biological relevance of the structure. A crystal structure may include lattice contacts that stabilize a conformation that is not dominant in solution. A cryo-EM structure may capture a more native-like state but at lower resolution. Check the entry metadata for the construct used, the presence of mutations, the buffer conditions, and any ligands or cofactors that were added.

For example, a structure solved with a stabilizing mutation may not represent the wild-type protein accurately. A structure with a bound detergent molecule may have a different conformation in the membrane environment. These contextual details are recorded in the PDBx/mmCIF file and the entry page, and they should be part of your comparison.

### Step 5: Document the Selection and the Rejected Alternatives

Record the accession code of the chosen structure, the date of download, and the metrics from the comparison table. Also record why you rejected the other candidates. This documentation is essential for reproducibility and for defending your choice in a publication or a thesis. A reviewer may ask why you chose one structure over another, and your comparison table provides the evidence.

The [EMBL-EBI training materials](https://www.ebi.ac.uk/training) include practical exercises on comparing structural entries and interpreting validation reports. Working through these exercises can help you build the skills needed to apply this framework consistently.

### Common Failure Patterns in Structure Selection

The most frequent mistake is choosing a structure solely by resolution without checking completeness or experimental context. A second common failure is ignoring the validation report entirely and assuming that any deposited structure is reliable. A third failure is using a structure without recording the accession code and download date, which makes the analysis impossible to reproduce later.

Another failure pattern is selecting a structure that does not contain the ligand or cofactor needed for the analysis. The presence of a ligand is recorded in the PDBx/mmCIF file, and the validation report may indicate whether the ligand has good density. If you need a bound ligand for docking studies, verify that the ligand is present and well ordered before proceeding.

### When to Escalate to a Professional

If the comparison table shows conflicting evidence, such as one structure with excellent geometry but poor completeness and another with good completeness but marginal geometry, consult a structural biologist who can interpret the tradeoffs in the context of your specific question. Similarly, if the validation report flags serious problems in all available structures, you may need to consider whether any existing structure is suitable for your purpose or whether you need to generate new experimental data.

For depositions of your own structures, the wwPDB OneDep system provides validation checks during the deposition process. If you are unfamiliar with the deposition requirements, seek guidance from a colleague who has deposited structures before. The [Bioconductor project](https://bioconductor.org/) and the [nf-core documentation](https://nf-co.re/docs) offer reproducible workflow examples that can help you document your structure selection and downstream analysis in a transparent manner.

## Frequently Asked Questions

### What is the difference between PDBx/mmCIF and the legacy PDB format?

The legacy PDB format uses fixed-width columns to record atomic coordinates, which limits the size and complexity of the structures that can be represented. PDBx/mmCIF uses a tag-value syntax that has no practical size limits and supports detailed annotation. New depositions to the PDB are required to use PDBx/mmCIF, and the legacy format is only available for older entries.

### How do I know if a cryo-EM map is good enough for my analysis?

Check the reported resolution of the map and the validation metrics for the model-to-map fit. A map with a resolution better than 3 angstroms is generally suitable for building a detailed atomic model, while a map with a resolution worse than 4 angstroms is only useful for understanding the overall architecture. The validation report provides the relevant metrics.

### What is R-free and why is it important?

R-free is a measure of how well the atomic model explains the diffraction data that were set aside during refinement. It provides an independent check of the model and helps identify overfitting. A structure with a high R-free or a large gap between R-work and R-free may not be reliable.

### What does the MolProbity score tell me?

The MolProbity score combines several geometric checks into a single number. It reflects the quality of the bond lengths, bond angles, backbone conformations, and the absence of steric clashes. A lower MolProbity score indicates better geometry, and the score can be used to compare the quality of different structures.

### Why are some residues missing from a structure?

Residues may be missing because they are disordered and do not produce clear density in the experiment. This is common for flexible loops and terminal regions. The missing residues are not modeled, and the PDBx/mmCIF file records which residues are present.

### Can I use a structure from the PDB for docking studies?

You can use a PDB structure for docking studies, but you must check the quality of the structure and the completeness of the binding site. The structure should have good geometry, and the binding site should be well resolved. If the binding site is missing or poorly defined, the docking results will not be reliable.

### How do I cite a structure from the PDB or EMDB?

The PDB entry page provides the citation information for the original paper and the database entry. You should cite both the paper and the database entry in your publications. The citation format is provided on the entry page.

### What should I do if the validation report flags problems?

If the validation report flags serious problems, you should not use the structure without understanding the issues. Consult a structural biologist who can help you interpret the report and decide whether the structure is usable for your purposes. You may also consider looking for an alternative structure of the same molecule.

## Related Bioinformatics Guides

- [The Protein Data Bank (PDB): Structural Formats, Coordinates, and Archival Validation Standards](/knowledge/bioinformatics/protein-data-bank-formats-archival-validation)
- [The Protein Data Bank (PDB): Archival Standards, Structural Validation Metrics, and Bioinformatics Integration Protocols](/knowledge/bioinformatics/protein-data-bank-archival-validation)
- [Genomic Data Analysis Tools: A Comparative Guide for Researchers](/knowledge/bioinformatics/genomic-data-analysis-tools-a-comparative-guide-for-researchers)
- [Lipidomic Analysis: A Beginner's Guide to Workflows and Data Interpretation](/knowledge/bioinformatics/lipidomic-analysis-a-beginner-s-guide-to-workflows-and-data-interpretation)
- [Persistent Identifiers for Research Data: A Guide to Selection and Use](/knowledge/bioinformatics/persistent-identifiers-for-research-data-a-guide-to-selection-and-use)

## References and Further Reading

- [NCBI Data Resources](https://www.ncbi.nlm.nih.gov/). National Center for Biotechnology Information.
- [EMBL-EBI Training](https://www.ebi.ac.uk/training). European Bioinformatics Institute.
- [Bioconductor](https://bioconductor.org/). Bioconductor Project.
- [Galaxy Training Network](https://training.galaxyproject.org/). Galaxy Project.
- [nf-core Documentation](https://nf-co.re/docs). nf-core.
- [The Carpentries Lessons](https://carpentries.org/lessons). The Carpentries.
- [The Eighth Edition AJCC Cancer Staging Manual: Continuing to build a bridge from a population-based to a more "personalized" approach to cancer staging.](https://pubmed.ncbi.nlm.nih.gov/28094848). CA: a cancer journal for clinicians, 2017.
- [A draft human pangenome reference.](https://pubmed.ncbi.nlm.nih.gov/37165242). Nature, 2023.
- [The data universe of structural biology.](https://pubmed.ncbi.nlm.nih.gov/32695409). IUCrJ, 2020.
- [QUAST: quality assessment tool for genome assemblies.](https://pubmed.ncbi.nlm.nih.gov/23422339). Bioinformatics (Oxford, England), 2013.
- [Advancing Molecular Dynamics: Toward Standardization, Integration, and Data Accessibility in Structural Biology.](https://pubmed.ncbi.nlm.nih.gov/38418288). The journal of physical chemistry. B, 2024.

> This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.