MetaboAnalyst

Spatial Data Bridge

Bring the exports of other tools into the two tables used by the spatial analysis modules (a spot x feature intensity table and a spot metadata table with x, y and a region column), derive regions of interest when an export has none, merge several sections into one data set, or pair a section with a bulk LC-MS table of the same tissue. Every utility ends on a preview page from which the converted tables can be downloaded or sent straight to the Spatial MS Peaks or Spatial Compound Table module. Raw formats (imzML, Seurat / SpaMTP objects) are converted with the MetaboAnalystR package on your computer. Size policy: each upload is limited to 50 MB, which covers a typical peak-picked section or a few sections at once; larger studies are merged in batches here, or processed end to end with the R package (same functions, no limit).

Format

Choose the tool the tables were exported from. Delimiters (comma, semicolon, tab) and comment lines are detected.

Files

Feature-intensity table:

Choose

The table export of SCiLS Lab: one row per spot with the spot index, region and x / y columns followed by one column per m/z. The region column becomes the ROI.

Section files

Choose

Select all files at once (Ctrl/Cmd-click), or one ZIP containing them. 50 MB in total.

Two files per section, in the standard format, named <section>_intensity.csv and <section>_meta.csv (for example brain1_intensity.csv + brain1_meta.csv). The section name is taken from the common prefix. Any number of sections; to go beyond the size limit, merge in batches and merge the results again.

Options

m/z matching tolerance (ppm):

Features are matched by name, or by m/z within the tolerance for peak tables; only features present in every section are kept. Sections are placed side by side and a Section column is added to the metadata, so the pseudobulk comparison can use either the section or the region labels.

For a section that also has a bulk LC-MS (or GC-MS) profile of the same tissue, typically with replicates and groups. The features of both platforms are matched, the matched spatial features take the LC-MS identities (compound names, if the LC-MS table has them) so that the section can continue in the Spatial Compound Table module, and the LC-MS statistics are reported next to the regional effect of each feature.

Spatial section

Intensity table:	Choose
Spatial metadata:	Choose

The two standard tables (spot x feature intensities, m/z or compound names; Spot_ID, x, y and a region column). A region column gives the regional effect size of each matched feature.

LC-MS table

Peak / compound table:

Choose

MetaboAnalyst statistics format: samples in rows with a Label column (or samples in columns with a Label row); features as compound names, or m/z (plain, mz__rt, mz/rt, M123T45). Two groups give a Welch t-test, more give an ANOVA; without labels only the abundance concordance is reported.

Matching

Ion mode:

Tolerance (ppm):

m/z are matched through the common adducts of the mode (e.g. a [M+K]+ ion on the MALDI side and [M+H]+ in LC-MS); compound names are matched exactly, or against m/z through their exact masses in the compound database. One partner per feature (smallest mass error).

imzML / ibd files (Cardinal) and Seurat or SpaMTP objects are converted on your computer with the MetaboAnalystR package, which writes the two standard tables for upload here:

library(MetaboAnalystR)
ReadImzML2Spatial("section.imzML", outPrefix = "section")      # Cardinal is required
ReadSeurat2Spatial("spamtp_object.rds", assay = "SPM", outPrefix = "section")
# writes section_intensity.csv and section_meta.csv

Both functions accept a region column of the object metadata (roiCol) and, for imzML, an optional peak-picking tolerance. See the package documentation for the arguments.