MetaLib: Shared Reference Library Cache

📋 On this page

Overview
How the Cache Works
What Each Plugin Stores There
A-SIMA's Metabolite Library
Cross-Referencing Between Libraries
Scope & Limitations

Overview

MetaLib isn't a plugin you open from a menu — it's POKY's shared reference-compound library, plus the local cache directory (also named MetaLib) that several plugins use to keep large datasets out of the POKY package itself. Instead of shipping megabytes of BMRB-derived spectra and shift tables with every install, a plugin downloads what it needs from MetaLib the first time it's used and reuses the local copy after that.

⬇️ Download MetaLib

Three plugins currently use this convention: A-SIMA (its full BMRB-derived metabolite library), P-Mix (its compound library for mixture design), and LigLib (its lignin & cell-wall model-compound database). Any future plugin needing a large reference dataset can follow the same pattern rather than inventing a new one.

💡 Browsing MetaLib's compounds directly: launch A-SIMA (two-letter command aa), then switch it to User Based mode — that's where you can search and browse the full MetaLib compound library yourself, rather than only seeing compounds A-SIMA has automatically matched to your spectrum.

💡 Where the data actually lives: the authoritative, most detailed write-up of the metabolomics side of MetaLib (compound counts, KEGG/HMDB category tagging, biospecimen origins) is on the POKY features site. This page instead documents the shared mechanism — the download/cache pattern itself — as it applies across A-SIMA, P-Mix, and LigLib.

How the Cache Works

Each MetaLib-consuming plugin follows the same three steps:

Resolve a local cache directory — normally a MetaLib folder created next to the currently open POKY project, created on demand if it doesn't exist yet.
Download on first need — if the data the plugin wants isn't in the cache yet, it's fetched from MetaLib and written into the cache directory.
Reuse the cached copy thereafter — later launches read straight from the local copy with no network access, until the plugin is told to refresh (or the cached file is deleted/moved, forcing a fresh download).

Some plugins cache their whole library up front; others additionally fetch per-compound data lazily, only once that specific compound is actually opened — this keeps the initial download small even though the full hosted collection is large.

ℹ️ No project open yet? Most MetaLib consumers key the cache directory off the current project, so open (or create) a POKY project before using a MetaLib-backed plugin for the first time. LigLib additionally falls back to a MetaLib folder under the user's home directory if no project is open, so it can still be browsed standalone.

What Each Plugin Stores There

A-SIMA — its full BMRB-derived metabolite library: compound names, BMRB IDs, structures, and experimental/simulated peak lists. See A-SIMA's Metabolite Library below.
P-Mix — a compound library (name, BMRB ID, formula, molecular weight, and more) used to build candidate mixtures, plus per-compound peak-list data fetched only once a compound is added to a mixture.
LigLib — 532 lignin/cell-wall model compounds (2024-08 NMR database edition): trivial/systematic names, category, linkage code, full ¹³C/¹H shift tables per solvent, and (for a subset) SMILES and an A-SIMA BMRB-ID cross-reference.

A-SIMA's Metabolite Library

A-SIMA is MetaLib's original and largest consumer. Its library covers 3,295 compounds across 3,629 entry IDs (3,409 experimental, 220 theoretical), sourced from the Biological Magnetic Resonance Bank (BMRB):

2D-HSQC data: 1,155 entries (822 experimental, 333 simulated)
1D data: 3,320 entries, all simulated (creatable from BMRB NMR-STAR files directly inside A-SIMA)
Categorization: 1,368 compounds are color-coded by class — acids, bases/neutrals, alcohols, amino acids, nucleosides/nucleotides/neurotransmitters, steroids, and plants/other — each also tagged with tier-1 KEGG pathway categories and biospecimen origin (blood, urine, cerebrospinal fluid, etc.)

⚠️ A-SIMA doesn't yet have a dedicated page in this manual site — for the full metabolomics-focused write-up (category color key, KEGG/HMDB details, a downloadable ZIP of the entire library), see the POKY features site.

Cross-Referencing Between Libraries

Because A-SIMA's library and LigLib's lignin database are both ultimately BMRB-derived, a compound can legitimately appear in both. LigLib cross-references A-SIMA's library by exact BMRB ID: 248 of LigLib's 532 compounds are also present in A-SIMA's library, letting you check a LigLib match against A-SIMA's independent peak-list/shift data for the same compound. See LigLib's Shift Table Viewer section for where that BMRB ID is shown.

Scope & Limitations

MetaLib-backed plugins need internet access the first time they're used (or the first time a not-yet-cached compound is opened); after that, cached files are read locally.
Each plugin defines its own file schema under MetaLib — there is no single unified "MetaLib format," only a shared hosting/caching convention.
Cached files aren't auto-refreshed. If a hosted library is updated, delete the corresponding file from the local MetaLib folder (or use the plugin's refresh option, where offered) to force a re-download.
Cache location is normally tied to the current POKY project directory, so the same machine can end up with multiple independent MetaLib caches across different projects.