Spectral Database Schema Validation

Spectral Database Schema Validation

Summary

Validate and reformat mass spectrometry transition data into the EISA-EXPOSOME schema required for suspect chemical screening. This skill ensures database files conform to the exact column structure and data types needed for high-throughput peak extraction and chemical annotation.

When to use

When you have compiled raw mass spectrometry transition data (precursor m/z, product m/z, intensity, retention time, compound IDs) from experiments or external sources, and need to prepare it for ingestion into the EISA-EXPOSOME R Shiny platform for suspect screening. Apply this skill before attempting any chemical annotation or peak extraction filtering.

When NOT to use

• Your database is already in EISA-EXPOSOME format and has been validated—use this skill only for initial preparation or reformatting.
• You are performing peak extraction or filtering on raw spectra—this skill validates the database, not the analytical workflow.
• Your input is already a feature table or chromatogram matrix; this skill is for transition-list database curation, not spectral processing.

Inputs

• Raw mass spectrometry transition data (CSV, XLSX, or tab-delimited text)
• Reference T3DB database file (.xlsx format) for validation cross-reference
• List of compound identifiers and expected transition pairs

Outputs

• Validated spectral database file (.xlsx or .csv) conforming to EISA-EXPOSOME schema
• Validation report confirming presence, formatting, and consistency of NAME, PrecursorMZ, ProductMZ, Intensity, RT, and ID columns
• Verified file integrity confirmation

How to apply

Obtain raw transition data in any tabular format (CSV, XLSX, or delimited text). Reformat into a table with exactly six columns: NAME (compound name), PrecursorMZ (numeric, e.g. 142.0086), ProductMZ (numeric, e.g. 94.0046), Intensity (numeric, typically 0–100 relative units), RT (numeric retention time in minutes; optional but recommended), and ID (integer or string identifier). Cross-reference your entries against the reference T3DB database file (provided in .xlsx format) to verify column alignment and confirm that well-known compounds like Methamidophos appear with expected transition pairs (e.g., PrecursorMZ 142.0086 → ProductMZ 94.0046 at Intensity 100, RT 2.182). Export the validated table as .xlsx or .csv and spot-check file integrity by reloading it into R or a spreadsheet editor to confirm no corruption or encoding errors occurred.

Related tools

• R Shiny (Interactive platform for visualization and filtering of validated spectral database results; runs the EISA-EXPOSOME interface after database validation is complete) — https://github.com/Lab-XUE/EISA-EXPOSOME
• T3DB (Reference database provided in .xlsx format; used as gold-standard cross-reference during column alignment and entry consistency checks) — https://github.com/Lab-XUE/EISA-EXPOSOME

Examples

# Pseudo-code R workflow using tidyverse and readxl; load raw transitions, rename/reorder to EISA schema, cross-check against T3DB, export:
raw_db <- read.csv('transitions_raw.csv')
eia_schema <- raw_db %>% select(NAME = compound_name, PrecursorMZ = prec_mz, ProductMZ = prod_mz, Intensity = rel_int, RT = rt_min, ID = cpd_id)
T3DB_ref <- readxl::read_excel('T3DB.xlsx')
validated <- left_join(eia_schema, T3DB_ref %>% filter(NAME == 'Methamidophos'), by = 'NAME') %>% filter(!is.na(validated_flag))
writexl::write_xlsx(validated, 'database_validated.xlsx')

Evaluation signals

• All six required columns (NAME, PrecursorMZ, ProductMZ, Intensity, RT, ID) are present and correctly named (case-sensitive).
• PrecursorMZ and ProductMZ values are numeric and match expected mass-to-charge ratios (e.g., Methamidophos 142.0086 and 94.0046, respectively).
• Reference compound entries (e.g., Methamidophos with expected transition pairs and RT 2.182) match T3DB entries exactly.
• Intensity values are numeric and typically in the range 0–100 (relative units); no text or null values in intensity columns.
• File loads without encoding or corruption errors; reloaded data shows identical row counts and value distributions as the exported file.

Limitations

• RT (retention time) is optional but strongly recommended; omitting it may reduce specificity in peak extraction workflows.
• The schema does not accommodate additional metadata columns (e.g., collision energy, instrument type, adduct type); extraneous columns should be dropped before export to avoid parser errors in EISA-EXPOSOME.
• Validation depends on availability of reference entries in T3DB; novel or rare compounds may lack cross-reference anchors.
• This skill validates schema and format only; it does not assess the scientific quality, accuracy, or completeness of transition data itself.

Evidence

• [readme] Database schema requirement: "If you are building your own database, your file (.xlsx /.csv) must contain the following columns:|NAME|PrecursorMZ|ProductMZ|Intensity|RT|ID|, RT is not essential."
• [other] Workflow steps for database preparation: "1. Obtain or prepare raw mass spectrometry transition data (precursor m/z, product m/z, intensity, retention time, and compound identifiers). 2. Format the data into a tabular structure with required"
• [other] Cross-reference and validation procedure: "3. Cross-reference entries against the compiled T3DB database file (provided in .xlsx format) to ensure column alignment and data consistency. 4. Validate presence and correct formatting of"
• [readme] Reference example entry: "|Methamidophos|142.0086|94.0046|100|2.182|1|"
• [other] Export and file integrity step: "5. Export the validated database as .xlsx or .csv format and verify file integrity."
• [readme] T3DB database availability: "We also provide the compiled T3DB database file in .xlsx format."