Phenotypic and clinical record comparison

Pheno-Ranker

Compare cohorts and rank target records from structured phenotypic and clinical data.

Pheno-Ranker compares Beacon v2 Models, Phenopackets v2, and configured JSON or YAML records using Hamming distance or Jaccard similarity. The command-line interface is the primary interface; CSV preparation and a web application are available as companion workflows.

Quickstart Supported Inputs Publication

Structured phenotypic and clinical records are compared by Pheno-Ranker to produce rankings and matrices — Structured records are compared to produce target rankings or cohort matrices.

Current scope

Supported operations

Available outputs and statistics depend on the selected mode and similarity metric. Input-specific requirements are documented with each supported format.

Cohort mode

Compute all-vs-all comparisons

Calculate pairwise Hamming distances or Jaccard similarities and write dense or sparse matrices.

Patient mode

Rank records against a target

Compare target records with reference cohorts and report ranked matches with overlap and completeness statistics.

Structured inputs

Use native or configured records

Read BFF and PXF directly, configure generic JSON or YAML, or prepare CSV data with the included import utility.

Result export

Inspect and reuse outputs

Export rankings, matrices, coverage information, intermediate records, and thresholded graphs for downstream analysis.