While guidance on how to submit to ArrayExpress or the BioImage Archive is available in the respective documentation pages, we are working to establish a shared metadata standard for spatial transcriptomics experiments.
Below are our recommendations for the metadata to include in your submissions in order to facilitate interpretation and reuse of your data.
If you have feedback about the metadata please reach out to us: spatial-transcriptomics@ebi.ac.uk
| Field | Definition | Example | Imaging-based ST | Sequencing-based ST |
|---|---|---|---|---|
| Sample ID | A unique identifier assigned to a biological specimen used in the imaging experiment | liver_01 | required | required |
| Slide ID | A unique identifier for the glass slide or cartridge on which the tissue section was mounted and processed | BSMC_01 | required | required |
| Position on slide | Describe where the capture area is located on the slide. This is required if the slide contains multiple capture areas, for example for visium | Area A1; number 1 | optional | optional |
| Sample preservation method | The method used to preserve the tissue prior to processing (e.g., fresh frozen, FFPE) | fresh frozen | required | required |
| Organism name | The scientific name of the source organism (e.g., Homo sapiens, Mus musculus) | Homo sapiens (NCBITaxon:9606) | required | required |
| Time from collection to preservation | Elapsed time between tissue excision/collection and preservation, reported in hours | 1h | optional | optional |
| Section thickness | The thickness of the tissue section mounted on the slide, reported in micrometers (µm) | 5 µm | required | required |
| Protocol for fixing section | Detailed description or reference to the fixation protocol used (fixative type, duration, temperature, buffer) | Tissues were fixed in 10% neutral-buffered formalin (VWR) for 24 hours at 4 ºC before being transferred to 70% ethanol at 4 ºC. Samples were then embedded in paraffin blocks and sectioned | required | required |
| Organ | Tissue origin | kidney (UBERON:0002113) | required | required |
| Anatomical region | The anatomical location from which the tissue was obtained, preferably using standardized anatomical ontologies. | Inner medulla of kidney (UBERON:0001294) | required | required |
| Donor health status at sampling | Relevant disease information from the donor at the time of sample collection | healthy, adenocarcinoma | required | required |
| Sample disease status | Disease information for the specific sample irrespective of the donor health status | benign, adenocarcinoma | optional | optional |
| Additional sample description | Any additional information about the sample that would be useful for reanalysis. For example can include: longitudinal sample collection, graft information, disease severity or grade | This sample is the second sample from a timeseries of samples collected from the donor, (2/5) samples collected 3 months after sample 1. | optional | optional |
| Sex | Biological sex of the donor organism | male | required | required |
| Age | Age of the donor at time of sampling | 35 | conditionally required (age or developmental stage) | conditionally required (age or developmental stage) |
| Developmental stage | Developmental stage for animal models | blastula | conditionally required (age or developmental stage) | conditionally required (age or developmental stage) |
| Sample collection protocol | Detailed description of the sample collection protocol, including the clinical procedure used to obtain the tissue (core biopsy, surgical resection, needle biopsy, collecting specimen from organ postmortem, etc) | Adult tissue was obtained from surgical remnants of heathy skin taken for reconstructive surgery. All samples were collected in Medium 154 (Gibco) with 1× Antibiotic-Antimycotic (Gibco) at 4 °C until dissociation. | optional | optional |
| Field | Definition | Example | Imaging-based ST | Sequencing-based ST |
|---|---|---|---|---|
| Sample preparation protocol | Sample preparation protocol which includes slicing, dewaxing, de-crosslinking, permeabilization method and time | Frozen samples were cryosectioned at − 10 °C (Thermo Cryostar). Tissue slice were placed on chilled Visium Spatial Gene Expression Slides (catalog no. 2000233, 10x Genomics), and stuck firmly by warming the back of the slide. Tissue sections were then fixed in chilled methanol and stained according to the Visium Spatial Gene Expression User Guide (catalog no. CG000239 Rev A, 10x Genomics). For gene expression samples, tissue was permeabilized for 18 min, which was selected as the optimal time based on tissue optimization time-course experiments. | required | required |
| Histological classification | General pathologist-based tissue classification | necrotic, hypoxic, vascularised, infiltrated...; | optional | optional |
| Permeabilization time | Permeabilization time and time units | 15 minutes | optional | required |
| Field | Definition | Example | Imaging-based ST | Sequencing-based ST |
|---|---|---|---|---|
| Imaging protocol | Imaging protocol which includes microscopy set up, instrument, magnification, and microscopy technique | Brightfield histology images were taken using a 10× objective (Plan APO) on a Nikon Eclipse Ti2-E (13,332×13,332 pixels2 for GEX) | optional | optional |
| Technology used | Technology used for the experiment | 10x Xenium, 10x Visium, MERfish, | required | required |
| Custom panel journal reference | Citation to the peer-reviewed publication describing or validating the custom gene panel | optional | ||
| Custom panel reference code | Vendor- or lab-assigned identifier or catalog code for the custom probe panel | optional | ||
| Segmentation signal | Imaging signals used to generate segmentations, including antibodies, stains, and channels | cell boundary stain is a combination of ATP1A1/CD45/E-Cadherin, RNA stain is 18S and nuclear stain is DAPI | required | optional |
| Physical pixel size x | Physical size that each pixel/voxel represent along the x-axis in um | 0.2125 | required | optional |
| Physical pixel size y | Physical size that each pixel/voxel represent along the y-axis in um | 0.2125 | required | optional |
| Physical pixel size z | Physical size that each pixel/voxel represent along the z-axis in um | 0.2125 | optional | optional |
| Instrument model | Type of instrument used for experiment | Xenium Analyzer | optional | optional |
| Chemistry version | Name of chemistry and version used | xenium v1 | required | required |
| Field | Definition | Example | Imaging-based ST | Sequencing-based ST |
|---|---|---|---|---|
| QC and filtering | Description of quality control steps and filtering thresholds applied to the data | optional | optional | |
| Version of analysis software | Version of the downstream analysis software used for data interpretation | xenium-3.2.0.7 | required | required |
| Method of annotating images | Description of how image features were annotated (manual, semi-automated, or automated) | semi-automated | required | optional |
| Annotation pair | Pair of source image and annotation mask | high_res_image.tiff cell_mask.tiff | required | optional |
| Image correlation between H&E and fluorescence | Method used to correlate images from different modalities, including fiducials or computational registration | Manual overlay | required conditionally | required conditionally |
| Segmentation method | Description of how the segmentations were created. Crowdsourced or expertly annotated. Software used and protocols used for consensus and quality assurance. Include key algorithmic parameters used for segmentation | The multimodal cell segmentation algorithm uses custom deep learning models trained on Xenium data. The new algorithm has three methods to segment cells. The final segmentation result for each cell is prioritized in this order: by boundary stain, by expansion from nucleus to interior RNA stain edge, or by isotropic nuclear expansion. For cases where cells that do not have boundary or interior stains, segment cells with a nuclear (DAPI) expansion distance of 5 µm or until another cell boundary is encountered | required | optional |
| Field | Definition | Example | Imaging-based ST | Sequencing-based ST |
|---|---|---|---|---|
| High resolution image | Raw or minimally processed high-resolution microscopy image files | required | ||
| Transcript coordinates | Spatial x, y (and z if applicable) coordinates of each detected transcript | required | ||
| Cell-level segmentation mask or polygons | Segmentation masks or vector polygons defining cell or nuclear boundaries | required | ||
| Probe panel | Machine-readable list of probe sequences, gene targets, and barcodes | required | ||
| Sample-level segmentation mask or polygons | Segmentation masks or vector polygons defining sample boundaries and class labels for sample ID - required for multiplexed sampled | conditionally required (multiplexed samples) | ||
| Field of view regions | Spatial definitions and boundaries for each imaging field of view | optional |
| Field | Definition | Example | Imaging-based ST | Sequencing-based ST |
|---|---|---|---|---|
| Fastqs | Raw sequencing read files in FASTQ format. | liver_01_R1.fastq.gz liver_01_R2.fastq.gz | required | |
| High resolution image | Raw or minimally processed high-resolution microscopy image files | Slide1_liver01.tiff | required | |
| Processed cell count matrix | Gene-by-cell matrix of transcript counts | barcodes.tsv.gz, features.tsv.gz, matrix.mtx.gz; feature_matrix.h5 | required | |
| Barcode coordinates | File containing the list of barcodes and their spatial coordinates on the slide | barcode_mappings.parquet | optional | |
| Probe panel | Machine-readable list of probe sequences, gene targets, and barcodes | gene_panel.json | optional |
