Spectra and separations - NMR, MS and chromatography end to end

1 · Learn the move · Recipe pattern

PARALLEL TRACK - complete only if pl1 pointed you here. The spectra and separations pipeline runs on one fact: no model reads your vendor binary. A Thermo .raw, an Agilent .d, a Bruker fid - all unreadable. The model authors the conversion and summary scripts; it never receives the file. Convert with the standard tool: msconvert for MS vendor formats to mzML, nmrglue for NMR (the acqus parameter file is already plain text), your CDS export for chromatography peak tables. Then shrink: mzML is XML and enormous - it looks pasteable and must not be pasted. The reasoning slice: run metadata, a TIC summary, a peak table with integration parameters, and one control: a standard whose retention time and area you already know. JCAMP-DX blocks look like readable numbers; they are compressed ASCII - hand the header, never the block.

Author my spectra/separations pipeline as a recipe. My files: [vendor format, e.g. Thermo .raw / Agilent .d / Bruker fid directory] plus [CDS peak table export]. Steps: (1) conversion command using [msconvert / nmrglue] to an open format, exact flags shown; (2) a summary script that extracts run metadata, TIC or spectrum summary, and scan counts - never the full data into chat; (3) peak table read that carries the integration parameters alongside the areas; (4) a control check: my standard at [known RT / shift, known response] must appear within [tolerance] or the run is flagged; (5) a run log appending every step. Mark anything about my method you need as [NEEDS: ...].