Other Formats

About other formats

Pheno-Ranker reads BFF and PXF directly. Other clinical standards should first be harmonized to BFF with Convert-Pheno, while arbitrary categorical data can be prepared through generic JSON/YAML or CSV workflows.

GA4GH native

BFF

Beacon v2 individuals.json; no custom configuration required.

GA4GH native

PXF

Phenopackets v2 records; no custom configuration required.

Clinical harmonization

Convert to BFF

Prepare OMOP, openEHR, FHIR, REDCap, or CDISC data with Convert-Pheno.

Beyond GA4GH

CSV

Converted to JSON plus configuration with csv2pheno-ranker.

Beyond GA4GH

Generic JSON/YAML

Custom categorical records described by a configuration file.

Clinical standards

For OMOP-CDM, openEHR, FHIR R4, REDCap, and CDISC data, use Convert-Pheno to produce BFF individuals.json, then run Pheno-Ranker with its built-in BFF handling.

YAML / JSON

Pheno-Ranker a priori accepts as input "any" JSON or YAML array file. In order to do it, you'll need a configuration file. See a tutorial on how to use it here.

What do you mean by "any" JSON or YAML?

Pheno-Ranker can process deeply nested data structures. From v1.08 onward, arrays nested more than one level deep are canonicalized automatically from their meaningful content, so equivalent nested objects can match even if their order differs between records.

For first-level arrays in generic JSON, set format: JSON and declare indexed_terms. You can also add identity_paths to keep user-facing keys more interpretable, but from v1.08 onward generic JSON can infer default identities when those paths are absent. For deeper nested arrays, no extra identity path is required.

CSV

We developed a utility that converts CSV files to JSON and automatically creates the configuration file needed. See an example on how to use it here.