Converter for JSON schema to mmCIF for PDB

This repository implements a file converterfrom OSC-EM JSON to PDBx/mmCIF. The JSON schema is defined in OSCEM and the PDBx/mmCIF format in (PDBML Schema v50 ).

Running the converter and required inputs:

Option 1 You can download the executable from the releases and use it in terminal with the following flags:

• with --json specify path to json file that contains metadata
• with --dic specify path to the PDBx/mmCIF dictionary file
• with --conversions specify path to conversions table. This table includes correspondance in names between OSC-EM and PDBx
• with --level specify the json element name that contains metadata entries. For SciCat that is usually "scientificMetadata"
• with --append specify if the metadata should be added to existing mmCIF to later deposit it in PDB
• with --mmCIFfile specify the path to existing mmCIF file. Throws an error if --append is false and --mmCIFfile is not specified
• with --output specify the file to write the newly created mmCIF with metadata entries

Option 2 If using in another GO application, you can install this package with

go get github.com/osc-em/converter-OSCEM-to-mmCIF 

and then run it with some of the

    ...
    parser.EMDBconvert(scientificMetadata, metadataLevelName, conversionsCsvPath, mmCIFdictPath, outputPath) // or
    parser.PDBconvertFromPath(scientificMetadata, metadataLevelName, conversionsCsvPath, mmCIFdictPath, outputPath, mmCIFInputPath)

The first argument in both functions is a scientificMetadata of type map[string]any that expects an unmarshalled JSON.

Here no checks against OSCEM schema are implemented, it is assumed metadata validates against this schema. Validation might still be added in the next releases.

Checks against mmCIF

Converter explicitly parser through the PDBx definitions to extract as much data as possible. This allows for

• administrative categories sorted within mmCIF ( such as author information, grant, etc)
• em-related categories are sorted randomly, as there is no definitive sorting in PDB team as well
• file ends with information on atoms
• units in OSCEM definition are compared to PDBx ( converter for units will be implemented )
• numeric values are checked to be within a range allowed by PDBs
• values are checked to be within a list of attributes allowed by PDBx. This is additionally enhanced to match via regular expressions or other applicable logic.

Behavior when category and data item is present in both metadata and provided mmCIF.

• There will be a logging message for this category
• In case it is a single instance in metadata (not an array of properties) and also a single instance in mmCIF (not within "loop") the values for respective category are generally taken from metadata. However, if a certain data item is present in mmCIF but not in metadata, it will still be added to the newly generated mmCIF
• In cases, where there is a mismatch between number of instances: number of properties in JSON array, and number of entries within "loop" in mmCIF, only the information from metadata will be used in newly generated mmCIF
• In cases, where there is a match between number of instances, the values for respective category are generally taken from metadata again and data items present in mmCIF will be added. CAUTION: The order of items is not checked and will be appended as is!