Metadata Recommendations

Metadata describes the biological and technical context of your dataset. Filling it in accurately benefits you and the wider community: it enables cross-dataset comparisons, supports reproducibility, and makes your data reusable — including by your own future self. Datasets with complete metadata are also more discoverable when made public on METASPACE.

The METASPACE team is actively working on metadata standardization to align with community standards and improve interoperability across datasets. The more specific and complete your metadata is now, the easier the transition will be — and the more valuable your datasets will be to others.

This guide focuses on the fields that have the greatest impact on reproducibility and are most commonly incomplete: Sample Information, Sample Preparation, and Dataset Description.

General Rules

• Complete as many fields as possible, even optional ones.
• If a field genuinely does not apply, enter N/A — do not leave it blank.
• Be specific. Vague entries are harder to interpret and impossible to use for grouping or comparison.
• For inspiration on standardized terminology, consult the Experimental Factor Ontology (EFO),BRENDA Tissue Ontology (BTO) or NCBI Taxonomy when filling in organism, tissue, and condition fields. Using controlled vocabulary makes your data easier to find and compare.

Sample Information

Organism

Enter the full scientific name of the species (e.g., Mus musculus, Homo sapiens, Arabidopsis thaliana). Avoid common names — "mouse" is ambiguous across species and strains; Mus musculus is not. Refer to NCBI Taxonomy if you are unsure of the correct name.

Organism Part

Specify the tissue or organ. Be as specific as the experiment requires — "brain" is acceptable if the whole brain was sectioned, but "hippocampus" or "cortex" is more informative if a specific region was targeted. For reference, the BRENDA Tissue Ontology (BTO) provides standardized tissue terms across a wide range of species.

Condition

Common mistake

This field is frequently left vague or inconsistent, which makes it impossible to group or compare datasets later. It is also used as a grouping variable in differential analyses — inconsistent labels will produce incorrect groupings.

Describe the experimental condition or treatment the sample underwent. If you plan to compare multiple datasets, decide on condition labels before uploading and use them exactly — including capitalization and spacing.

❌ Avoid	✅ Better
treated	treated with 10 µM tamoxifen, 48h
sick	high-fat diet, 12 weeks
control	untreated, wild-type
KO	PTEN knockout, C57BL/6 background

For standardized condition terminology, refer to the Experimental Factor Ontology (EFO), which covers diseases, treatments, cell lines, and experimental variables across biomedical research.

Sample Growth Conditions (optional)

Common mistake

This field is often skipped entirely, even when culture or housing conditions are a key experimental variable.

Describe how the sample was grown, housed, or maintained prior to collection. Examples by sample type:

• Cell culture: growth medium, serum percentage, CO₂ level, passage number, confluency at harvest
• Animal model: housing conditions, diet, light/dark cycle, age and sex at time of collection
• Plant: growth medium, light cycle, temperature, developmental stage

This field is especially important for in vitro experiments, where culture conditions can substantially affect the metabolome.

Sample Preparation

These fields are primarily relevant for MALDI-based tissue experiments. Enter N/A for any field that does not apply to your setup. Even so, do not skip them without consideration — incomplete sample preparation metadata is one of the main barriers to reproducing IMS experiments.

Sample Stabilization

Common mistake

Users often skip this field or enter the tissue type rather than the preservation method.

Describe how the tissue was preserved after collection. Common entries:

• Snap-frozen in liquid nitrogen immediately after dissection
• Fresh-frozen without fixation, stored at −80°C
• Formalin-fixed paraffin-embedded (FFPE)
• OCT-embedded, snap-frozen

The preservation method directly affects metabolite integrity and is essential information for anyone attempting to replicate your experiment.

Tissue Modification

Describe any physical or chemical alterations made to the tissue before matrix application. Examples:

• Cryosectioned at 12 µm, thaw-mounted on ITO-coated glass slides
• Deparaffinized prior to analysis (for FFPE sections)
• On-tissue trypsin digestion prior to matrix application
• No modification

MALDI Matrix

Enter the matrix compound using its common abbreviation and full name (e.g., DHB (2,5-dihydroxybenzoic acid), 9AA (9-aminoacridine), DAN (1,5-diaminonaphthalene)). Matrix choice is a key determinant of which compound classes are detected.

MALDI Matrix Application

Describe the method and key parameters of matrix deposition. Include the device name and relevant settings where known. Examples:

• Automated pneumatic sprayer (HTX TM-Sprayer), 8 passes, 40°C
• Sublimation at 160°C, 15 min
• Manual spray using a TLC sprayer, 6 passes

Solvent

Enter the full solvent composition used to prepare the matrix solution, including concentrations and any additives (e.g., 30 mM DHB in 70% methanol / 30% water with 0.1% TFA). Solvent composition affects matrix crystallization and ionization efficiency and is necessary for replication.

Dataset Description

TIP

The description field is the most flexible and most underused field in the upload form. Use it to include your full sample preparation protocol.

A thorough dataset description makes your data independently interpretable without requiring contact with you or your lab. At a minimum, include:

• Experimental aim — what biological question the dataset addresses
• Sample collection — when and how the tissue or cells were collected, including any animal model details or cell line information
• Full preparation protocol — step-by-step, from tissue collection to matrix application, including reagents, equipment, and timings
• Acquisition parameters — instrument settings not captured elsewhere, such as laser frequency, mass range, and number of laser shots per pixel
• Any deviations from standard protocols or known issues with the dataset

Example Description

Mouse liver tissue was collected from 16-week-old male C57BL/6J mice fed a high-fat diet (60% kcal from fat) ad libitum from 8 weeks of age. Tissue was snap-frozen in liquid nitrogen immediately after dissection and stored at −80°C. Cryosections (10 µm) were thaw-mounted onto ITO-coated glass slides and dried in a vacuum desiccator for 30 min. DHB matrix (30 mM in 70% MeOH / 30% H₂O + 0.1% TFA) was applied using an HTX TM-Sprayer (8 passes, 40°C, 0.1 mL/min flow rate). Data were acquired on a Thermo Orbitrap Elite in positive ion mode at 140,000 resolving power (m/z 200), mass range m/z 200–1000, 50 µm pixel size, 10 laser shots per pixel.

This level of detail is achievable in a few sentences and makes a substantial difference to reproducibility. If your lab maintains a standard operating procedure (SOP) for sample preparation, paste the relevant sections here directly.