LigLib: Lignin NMR Model Compound Library

📋 On this page

Overview
Data Source & Coverage
Browse Library Tab
Shift Table Viewer Tab
Peak Matching & Assignment Tab
Scope & Limitations

Overview

LigLib bundles the NMR Database of Lignin and Cell Wall Model Compounds (S. Ralph, J. Ralph & F. Lu; US Forest Products Laboratory / USDA-ARS Dairy Forage Research Center / Great Lakes Bioenergy Research Center, 2024-08 edition, public domain) directly inside POKY, so lignin/plant-cell-wall NMR spectra can be browsed against known monomer, dimer, trimer and tetramer model compounds without leaving the application.

💡 Where to find it: LigLib registers as a single menu command (accelerator LG, under the Small Molecule Analysis menu) that opens a three-tab dialog: Browse Library, Shift Table Viewer, and Peak Matching & Assignment.

Data Source & Coverage

LigLib covers 532 compounds from the 2024-08 edition of the NMR Database of Lignin and Cell Wall Model Compounds: compounds 1–340 (Vol. 1), 1001–1030 (Vol. 2; compound 1031 does not exist in the source), 2001–2082 (Vol. 3), and 3001–3080 (Vol. 4) — monomers, β-O-4/β-5/β-β/5-5/β-1/4-O-5 dimers, trimers and tetramers, benzodioxane and dibenzodioxocin structures, resinols, Hibbert's ketones, thioacidolysis products, and several Morreel-series oligolignols — transcribed from the database's PDF data pages. Each compound entry carries:

Trivial and systematic names, structural category, linkage-code notation (e.g. G-b-G, S-(8-5)-G) and erythro/threo isomer label where given
Full ¹³C and ¹H shift tables (multiplicity and J-couplings where reported), per solvent — acetone-d₆, CDCl₃, and/or DMSO-d₆
Source/notes fields carried over from the original data sheets (sample ID, contributor, literature reference)
A best-effort-matched SMILES string for 412 of the 532 compounds (via A-SIMA BMRB-ID cross-reference, PubChem name/CAS lookup, and NIH CACTUS name-to-structure resolution) — see the note in Shift Table Viewer

ℹ️ The compound library is downloaded, not bundled: LigLib does not ship its compound data inside the POKY package. On first use it downloads it from MetaLib and caches it locally alongside the current project — the same shared convention used by P-Mix and A-SIMA. Subsequent launches reuse the cached copy; delete the cached file (or restart with a fresh project directory) to force a re-download after the hosted file is updated.

Browse Library Tab

A searchable table of every bundled compound (ID, trivial name, category, linkage code, isomer, and which solvents have data). The search box matches against name, category, linkage code, source, and notes text; Category and Solvent dropdowns narrow the list further. Double-clicking a row (or selecting one and clicking "View Shift Table") jumps to that compound in the Shift Table Viewer tab.

Shift Table Viewer Tab

Shows a chosen compound's full metadata (systematic name, category, linkage code, isomer, source, reference, notes) alongside two tables — one for every ¹³C shift row, one for every ¹H shift row (with multiplicity and J-coupling) — each row tagged by ring letter and solvent. A "Plot solvent" dropdown drives an embedded scatter plot of the compound's paired (¹H,¹³C) correlations (axes inverted, high-to-low ppm, in the usual HSQC convention) with each point labeled by its atom name — a quick visual fingerprint to compare against an experimental HSQC.

ℹ️ Structure images: each compound entry carries a SMILES field, and LigLib renders a 2D structure from it via RDKit when both a SMILES string and RDKit are available. 412 of the 532 compounds currently have a SMILES string, matched in three passes: (1) exact BMRB-ID cross-reference against A-SIMA's reference library where the same compound is present there, (2) best-effort PubChem name/CAS lookup, and (3) NIH CACTUS name-to-structure resolution of well-formed systematic names for the remaining gap. Every match from passes 2–3 was cross-checked against this compound's own recorded category/systematic name before being accepted; ambiguous or internally-inconsistent hits (e.g. PubChem resolving LigLib's own linkage-code shorthand like G-b1-G to an unrelated compound, or CACTUS silently mis-parsing "…diacetate" suffixes as disconnected acetic acid) were discarded rather than risk showing a wrong structure. The remaining 120 compounds are left blank; only text-based shift tables and the correlation plot are shown for those. See Scope & Limitations.

ℹ️ A-SIMA cross-reference: compounds that also appear in A-SIMA's BMRB-derived compound library carry that compound's BMRB ID (e.g. bmse010330) in a bmrb_id field — useful for cross-checking a LigLib match against A-SIMA's independent peak-list/shift data for the same compound.

Peak Matching & Assignment Tab

Pick an open spectrum from the spectrum menu; LigLib inspects its nuclei to auto-identify the ¹H and/or ¹³C axis. Set ¹H/¹³C match tolerances (defaults 0.03 and 0.5 ppm) and optional Category/Solvent filters, then click Scan Peaks for Matches: every peak in the spectrum's peak list is compared against every library compound's (¹H,¹³C) correlations (or single-axis shifts, for a 1D spectrum), and the best five matches per peak (ranked by tolerance-normalized combined distance) are listed with the peak position, matched compound, atom/ring label, solvent, and Δ¹H/Δ¹³C.

Selecting a match row pre-fills a suggested Group name (e.g. C3001A) and Atom name (e.g. Abeta) from the matched compound/atom; both are editable before applying. Go to Selected Peak (or double-clicking the row) centers and raises the spectrum view on that peak; Apply Assignment to Selected Peak confirms and then calls the peak's assignment on its ¹H and/or ¹³C axis, refreshing its on-screen assignment label.

⚠️ Matches are chemical-shift proximity only, not a structural/statistical confidence score — always sanity-check a suggested match (correct ring pattern, multiplicity, expected linkage chemistry) before applying it as an assignment.

Deconvolve with REDEN manually launches POKY's REDEN peak-decomposition dialog (2D or 3D, chosen automatically from the currently selected spectrum view) — use it on ambiguous or overlapped peaks before trusting a match, then re-run Scan Peaks for Matches against the deconvolved peak list. This is a manual launch only; LigLib does not drive REDEN's fitting automatically.

Scope & Limitations

Covers compounds 1–340, 1001–1030, 2001–2082, and 3001–3080 of the 2024-08 database edition (532 total). Compound 1031 does not exist in the source PDF (Vol. 2 stops at 1030). Earlier/later database editions are not covered.
SMILES/2D structure images are only available for 412 of 532 compounds (see the note in Shift Table Viewer) — the remainder show text-based shift tables and the correlation plot only.
(¹H,¹³C) correlation pairs are built by matching same- labeled rows between a compound's separate ¹³C and ¹H tables in table order; a handful of source compounds with only partial side-chain assignments or ambiguous isomer-mixture shifts (noted in their "Notes" field) may pair imperfectly.
Peak matching scores each axis independently against its tolerance and sums the normalized distances — it is a nearest-neighbor lookup, not a fitted or machine-learned confidence model.
REDEN deconvolution is a manual, one-way launch from the Peak Matching tab — results must be re-scanned by hand; LigLib does not read REDEN's fit parameters automatically.