Source data and code for Figure 2 of ZT Scan DIA spectral data mining for metabolomics using MS-DIAL 5.6

This deposit holds the data and code behind Figs. 2b–2e of doi.org/10.1101/2025.08.20.671307. Those figures evaluate how the Q1-axis deconvolution (Q1Dec) and the retention-time deconvolution (RTDec) of MS-DIAL 5.6 affect spectral-library-based annotation in hydrophilic metabolomics data. The data are from NIST SRM 1950 human plasma, acquired in positive ion mode by DDA, SWATH-DIA and ZT Scan DIA, with three technical replicates each. The ZT Scan DIA data were processed at Q1-bin tolerances of ±0 to ±3 bins, with and without RTDec. Fig2_final.zip contains: - R/: three R scripts, one for Fig. 2b, one for Fig. 2c and one for Figs. 2d and 2e. - data/: the spectral similarity scores of the 13 acquisition and processing settings (inputs of Figs. 2b and 2c), and the MS-DIAL peak lists of the nine samples (inputs of Figs. 2d and 2e). Each score file combines the peak lists of the three replicates and holds one row per peak with an MS2 spectrum and library-match scores. - source_data/: the values as plotted in each panel. - figures/: the panels as drawn by the scripts (PDF). - README.md and MANIFEST.csv (the size and MD5 checksum of every file). All inputs are included. After unpacking, run the three scripts from the package root with R (tested with 4.5.1, tidyverse and patchwork); each takes under half a minute. README.md describes the files and their columns, the processing setting of each file, how each panel is computed, and the changes from the scripts used for the article. The reference spectra of the in-house library are not included. The data and code of Figs. 3 and 4 are deposited separately (https://doi.org/10.5281/zenodo.23692676). An earlier version of that record (https://doi.org/10.5281/zenodo.20812132) also held the data of Fig. 2, which are now deposited here.