Skip to content

Molecular Analysis

Overview

The Molecular Analysis tab is the primary entry point for ¹³C / DEPT spectral processing and natural product database screening. Main capabilities:

  • Spectral input and peak picking: Load Bruker pdata for ¹³C, DEPT90, and DEPT135; detect peaks with adjustable intensity thresholds; merge peaks into a unified peak table (NMR-1D.csv).
  • Signal cleanup: Remove solvent and impurity signals by carbon type (CH₃, CH₂, CH, C).
  • Manual or file-based peaks: Optionally skip preprocessing by pasting ppm, q/t/d/s lines.
  • Database search: Load built-in or custom .npz libraries; apply CNF / CTNF / MW filters; rank hits with CSS, AAS, FPS, or FPAACS (default Top 80).
  • Results inspection: Table of candidates; Compound Details popup with structure, predicted vs experimental spectrum, export, and AI interpret.
  • Methods export: Copy methods text builds an English Methods paragraph from current GUI parameters (including Fragment settings).

Interface Layout

The tab uses a left control panel (max width ~650 px, scrollable) and a right results panel: Molecular Analysis — main window Left panel

  • Collapsible Spectrum Preprocessing (collapsed by default): spectral paths, peak thresholds, solvent/impurity removal.
  • Collapsible Database Analysis: manual peaks, database selection, filters, evaluator, Start Analysis / Copy methods text / AI: Top 5 →.
  • Progress bar (visible during database analysis).

Right panel

  • NMR Spectrum Display: Matplotlib canvas + toolbar; toggle Peak Detection Plot / Merged Spectrum.
  • Peak Information (NMR-1D.csv): chemical shift, carbon type, intensity.
  • Analysis Results: five-column table; click row for details; context menu for AI actions.

Workflow

① Set spectral data folders (or skip with manual peaks)

  • Expand Spectrum Preprocessing → Spectral Data Input.
  • Set three Bruker processed data directories (typically ...\pdata\1):
Field Description Example (demo)
13C_NMR Broadband ¹³C spectrum test\NMRsample\4\pdata\1
DEPT90 DEPT-90 test\NMRsample\8\pdata\1
DEPT135 DEPT-135 (positive/negative) test\NMRsample\6\pdata\1
  • Use the folder icon to browse. Data are read with nmrglue (ng.bruker.read_pdata).

Skip this step

If you already have a peak list, fill Manual peak input under Database Analysis (step ④) and go directly to ⑥ after loading the database.

② Run peak detection and merge

  • Expand Peak Detection Parameters.
Parameter Default Description
C 3000000 ¹³C peak-picking threshold (instrument-dependent)
C90 90000000 DEPT90 threshold
C135pos 50000000 DEPT135 positive threshold
C135neg -50000000 DEPT135 negative threshold
  • Click Submit. On success: right panel shows peak-detection view; Molecular Analysis — peak-detection
  • Click Merge. The tool sorts carbon types, reconstructs the combined spectrum, and writes GUI_result_files/NMR-1D.csv. Molecular Analysis — Merge

③ Optional: Solvent and impurity removal

Solvent Removal (after Merge)

Control Options
Solvent Chloroform, Methanol, DMSO, Pyridine
Type CH3, CH2, CH, C
Submit Removes solvent peaks for the selected carbon class

Impurity Removal

Control Description
Type CH3, CH2, CH, C
Threshold Two factors; effective threshold = Threshold1 × Threshold2 (default second box 1e7)
Submit Requires Threshold1; uses absolute method

Each step updates NMR-1D.csv and the merged spectrum view.

④ Load database and set filters / evaluator

Database Selection

Dropdown Internal file Size (UI label)
Plant Database Database/Plant.npz 188,478 NPs
Human Database Database/Human.npz 217,347 NPs
Microbial Database Database/Microorganism.npz 36,427 NPs
Drug Database Database/Drug.npz —
All Database Database/AllDB.npz 605,735 NPs
  • Load Database or browse Other Database (.npz) → Load Other Database.
  • When loaded, Start Analysis is enabled.

Filter Parameters

Filter Parameter Default Description
CNF ✓ CNF bias 5 Carbon-count filter tolerance
CTNF ✓ CTNF bias 2 Carbon-type count filter (q/t/d/s)
MW MW list 300,400 Molecular weight filter when checked

Evaluator Parameters (choose one)

Evaluator Extra parameter Default Description
CSS CSS threshold 0.5 Chemical-shift similarity style score
AAS — — Average absolute deviation style
FPS — — Fingerprint similarity
FPAACS ✓ FPAACS weights 0.2,0.3,0.5 Weighted combination (default)

⑤ Start analysis and review results

  • Click Start Analysis. Progress runs in AnalysisThread; results saved to GUI_result_files/Result.csv.
  • Analysis Results columns: Compound ID, SMILES, Score, Chemical Shifts, Carbon Types. Molecular Analysis — Merge
  • Click a row → Compound Details: SMILES copy, Previous/Next, spectrum canvas Molecular Analysis — Merge
  • Optional: Copy methods text, AI: Top 5 → (switches to AI Assistant tab).

⑥ Output file structure

Under GUI_result_files/:

GUI_result_files/
  NMR-1D.csv      # After Merge / solvent / impurity (ppm, type, intensity)
  Result.csv      # After Start Analysis (Top-N hits)

Legacy: if NMR-1D.csv exists only in the app root, it is used when the folder above has no file.

FAQ

  • "Please select all required data folders" on Submit Set all three spectral paths before peak detection.

  • No peaks / peak detection failed Lower C, C90, or C135 thresholds; confirm paths are valid Bruker pdata and nmrglue is installed.

  • Start Analysis: no peak data Run Merge or fill Manual peak input with valid ppm, type lines.

  • Failed to load database Ensure Database/Plant.npz (etc.) exists relative to the application root. Custom libraries must match the format from Database Management.

  • "No compounds meet the search criteria" Widen CNF/CTNF/MW filters; check that experimental peak count and q/t/d/s pattern are compatible with the library.

  • Merge opens extra Matplotlib windows The GUI forces Qt5Agg and DrawWin="False"; if pop-ups appear, close other scripts using interactive pyplot.

  • Open / Save Project does nothing useful .vmp project files are placeholders; rely on Result.csv, exports, and screenshots.

Relation to Other Modules

  • Fragment Analysis: Requires Result.csv / non-empty result table and experimental peaks from this tab (file or manual input). Run masking attribution is enabled after successful analysis.
  • AI Assistant: Uses screening results, peaks, and optional fragment context for Top-N and diagnosis; Copy methods text is shared with this tab.
  • Database Management: Produces custom .npz files loaded here via Load Other Database.
  • Other Tool: Independent SMILES drawer; no link to screening results.