[Feature] Edit category names and change biomarker categories

[Jonseed]

Description

Currently there is no way to change category names or change which category a biomarker is in. For example, I have some biomarkers that imported as "Urine Amino Acids" but there was no urine test. They should be under an "Antioxidant" category instead. Maybe each detailed biomarker card could have a dropdown menu that lets you change which category that biomarker is in? I guess the problem is that some biomarkers can probably be categorized in multiple categories too. How should that be handled?

Another reason to be able to edit category names, I currently have one labeled "Apo A-1" that I'd like to rename "Apolipoprotein A1".

System Info

• Browser: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/144.0.0.0 Safari/537.36 Edg/144....
• Screen: 1920x1080
• Theme: dark
• AI Provider: ollama

[elkimek]

Implemented in v1.7.2 — you can now:

• Rename any category — click the category name in the header to edit it
• Change any category's icon — click the icon to open a built-in emoji picker with categorized emojis, search, and a reset button

Both work on standard and custom categories. The underlying data keys are unchanged, so existing exports and imports stay compatible.

Moving a marker between categories isn't supported yet — that would require remapping all historical data. If there's a specific use case for it, happy to hear more.

[Jonseed]

Thanks. I currently have three amino acid categories under OAT:

• Amino Acids & Lipids
• Urine Amino Metabolites
• Urine Amino Acids

I can rename the "urine" categories to remove that label, since there was no urine test (these markers are not exclusively urinary markers), but the biomarkers in these categories could probably also be consolidated into a single "Amino Acids" category.

Some of the biomarkers that got categorized there should probably be moved to a different category altogether, such as acetoacetic acid, which is a ketone body and part of energy metabolism (it is not an amino acid).

Sometimes the AI will just categorize markers wrong, or create duplicate categories, and it seems like there should be a way to correct for that, managing categories, moving markers between them, or even tagging markers with multiple categories or tags if they fall into multiple domains (as they do on my lab report).

[elkimek]

The rename part is done (v1.7.2) and you can also now rename individual markers (#64).

Moving markers between categories is on the radar but it's a bigger change — every data entry, reference override, and custom marker definition uses category.markerKey as its key, so moving a marker means remapping across all historical data. Not impossible, but needs careful handling to avoid data loss.

For now, renaming the categories ("Urine Amino Acids" → "Amino Acids") is the quick fix. For misclassified markers like acetoacetic acid, the best workaround is to delete the custom marker and re-import — the AI prompt can be nudged with a note in the chat before importing.

[Jonseed]

🚀 Proposal: Future‑Proof Biomarker Schema for Flexible Categorization

Below is a short, implementation‑agnostic outline of what a more robust biomarker data model could look like. The goal is to support:

• renaming categories
• consolidating categories
• moving biomarkers between categories
• tagging biomarkers with multiple domains
• preserving compatibility with existing data

1. Stable Internal Marker IDs

Each biomarker should have a permanent, category‑independent internal ID, for example:

markerId: "ms_000123"

This ID becomes the primary key referenced by:

• historical data entries
• reference range overrides
• custom marker definitions
• imports/exports

This removes the need for category.markerKey as a compound key.

2. Category as Editable Metadata

Categories become metadata, not part of the primary key. A biomarker could have:

categories: ["Amino Acids", "Energy Metabolism"]

This allows:

• renaming categories
• consolidating categories
• reassigning biomarkers to different categories
• multi‑category classification

All without breaking historical data.

3. Optional Tagging System

Beyond categories, biomarkers could support tags for finer‑grained grouping:

tags: ["ketone body", "OAT", "mitochondrial"]

Tags allow flexible UI filtering without forcing rigid category structures.

4. Migration Layer for Legacy Keys

To maintain compatibility with existing exports/imports, a simple migration layer could map:

oldKey: "Urine Amino Acids.acetoacetic_acid"
→ markerId: "ms_000123"

This ensures:

• no data loss
• old exports still import cleanly
• users can gradually transition to the new schema

5. UI Benefits

With this schema, the UI could support:

• drag‑and‑drop category reassignment
• multi‑category tagging
• category consolidation tools
• AI‑assisted reclassification
• cleaner navigation for mixed panels (CBC, chemistry, OAT, metabolomics, etc.)

6. Why This Matters

The current schema couples category to the primary key, which you have already noted makes moving markers difficult because it requires remapping all historical data. Decoupling identity from category removes this rigidity and aligns the data model with how biomarkers behave in real biochemistry: many belong to multiple pathways, and classification evolves over time.

[elkimek]

This is a well-thought-out proposal — the stable marker IDs + category-as-metadata approach is the right direction long-term.

That said, I have to be honest about capacity. I can fix bugs and polish things, but designing and implementing a schema migration of this scope is a significant effort. A PR with a complete solution would be much more effective than going back and forth on details in issues.

You seem well-equipped to work with the codebase. If you're up for it, I'd love to see a PR — even a draft/WIP to get the conversation going with actual code. CLAUDE.md covers the architecture, and the data flow section explains how category.markerKey is used throughout. Happy to review, test, and help polish anything you put together.

Same goes for some of your other issues that are specific to your use case and hard for me to test without your data — PRs are the fastest path there too.

[elkimek]

Category renaming already works (via categoryLabels/markerLabels overrides in the detail modal). But moving a marker to a different category is a structural limitation — v1 marker keys embed the category (biochemistry.glucose), so changing the category would break all existing data references, refOverrides, and chart history.

The v2 architecture (TypeScript/Svelte/Tauri rewrite) normalizes the data model so category is just a column on the marker definition. Reassigning a marker to a different category becomes a simple column update without breaking value history. Targeting this as part of the v2 data model redesign.

Leaving this open as a v2 milestone.

[fpietrosanti]

A biomarker coming from an exams should have a LOINC ID for Request and Response as a data structure.
Have you considered using those interoperable data format?
(I'm gonna build a biohacker thing, analyzing data structure, i think your software is a great piece of inspiration!)

[elkimek]

@fpietrosanti Thanks for the suggestion! LOINC mapping is a great idea for interoperability — I've opened #98 to track it.

The plan is to add an optional loinc field to each marker in the schema, then build a lookup table for the ~200 standard biomarkers. This would enable FHIR import/export down the line (Apple Health, MyChart, EHR systems).

Currently the AI parser maps PDF text to internal category.markerKey identifiers, but LOINC would add an unambiguous standard layer on top.

Good luck with your biohacker project!