# 03 Identifier Linking

## Quick Answer
Identifier linking is the backbone of EVd3x. Molecules are resolved to canonical IDs, then joined across targets, pathways, expression, disease, and interaction tables.

## What this does
Defines how records are linked and normalized across heterogeneous tables.

## Inputs
- miRNA query names and IDs
- Gene symbols and `Gene_ID`
- Protein IDs and STRING protein references
- Disease keys (`Disease_ID` preferred)

## Outputs
- Canonicalized node IDs in network elements
- Grouped disease associations
- Join-ready exports with IDs and display labels

## How calculated
Key mapping examples:
- miRNA: query name -> `miRBase_ID`
- Targets: `miRNA_ID` -> `Gene_ID`
- Proteins: `Corresponding_Gene_ID` -> `Protein_ID`
- Disease: grouped by `Disease_ID` with fallback name key only when ID is missing

## What to download
Use files that include both canonical IDs and display names to preserve reproducibility and readability.

## Known limits
When source data lacks canonical IDs, fallback grouping can still be ambiguous. Raw names are retained for traceability.