OpenPedCanExpress is a browser-based tool for exploring gene expression in pediatric tumors compared with normal tissue controls. Search a gene, compare its expression across tumor histologies and normal samples, and export publication-ready plots.
EGFR, TP53, or CD276) and press Enter.Configure Groups to choose which cohorts, histologies, molecular subtypes, and normal tissues to show.Box width), Height, Font size, and file type (PDF, JPG, PNG, or SVG).Export Plot to save the figure.The first load takes a few seconds while the database engine starts. After that, each gene search only downloads the rows for that gene.
PBTA, TARGET, GMKF), then normal controls (Pediatric Brain, Evo-devo, GTEx <40). Facet widths are proportional to the number of groups.log₂ scale option to see raw TPM. The y-axis is shared across all facets so groups can be compared directly.Open Configure Groups to choose what is plotted.
Cohorts. All cohorts (PBTA, TARGET, and GMKF) are shown by default, each in its own facet. Uncheck a cohort to hide it.Histologies. Tumor groups, listed under their cohort heading.Molecular subtypes. Select subtypes to plot those samples separately. Split by molecular subtype divides every tumor group by subtype.Control cohorts. Choose GTEx, Evo-devo, and pediatric normal brain groups.Click Apply to update the plot.
Choose a format (PDF, JPG, PNG, or SVG; raster images are 300 DPI) and click Export Plot. The export is drawn independently of the on-screen plot, so you can set:
Box width: pixels per group, which sets the figure width.Height of the figure.Font size: margins scale with it so labels are not clipped.Expression values are RSEM TPM, collapsed to gene symbol, from the
OpenPedCan project
(OpenPedCan v15 data release).
All data, tumor and control, was harmonized with the Gabriella Miller Kids First RNA-Seq workflow using the GENCODE v39 reference, so expression values are comparable across cohorts.
A preprocessing script (scripts/01-build-expression-parquet.R) converts them into Parquet files split by the first letter of the gene symbol (A–H, I–P, Q–Z). The app queries these files in the browser with DuckDB-WASM, so there is no server or database.
PBTA, TARGET, and GMKF cohorts.PBTA neuroblastoma specimens are excluded.| Control | Description | Data access |
|---|---|---|
| GTEx <40 | GTEx brain regions, donors aged 20–39 | dbGaP phs000424 |
| Pediatric Brain | Non-neoplastic pediatric tissue: Cerebellum, Cortex, Pituitary, Pons | BioProject PRJNA1019377 |
| Evo-devo | Developmental brain from 4 weeks post-conception through elderly | ArrayExpress E-MTAB-6814 |
The same control cohorts are used in TAPESTRY.
PBTA. Histology groups come from TAPESTRY, so colors and groups match across Rokita Lab tools.TARGET and GMKF. Samples are grouped by OpenPedCan cancer_group. Groups with 3 or fewer samples are not shown, and Ganglioneuroblastoma is combined with Neuroblastoma. The mapping is in plot-groups.tsv.| Term | Meaning |
|---|---|
| TPM | Transcripts per million, a within-sample normalized expression value |
| log2(TPM + 1) | TPM on a log scale; the +1 keeps zero values defined |
PBTA | Pediatric Brain Tumor Atlas cohort |
TARGET | Therapeutically Applicable Research to Generate Effective Treatments |
GMKF | Gabriella Miller Kids First |
| GTEx | Genotype-Tissue Expression project (normal tissue) |
| Independent specimen | One representative specimen per participant, so no participant is counted more than once |
| Plot group | The category a sample is plotted under (a histology or a normal tissue) |
OpenPedCanExpress is developed by the Rokita Lab. Please also cite OpenPedCan for the underlying data. Questions and bug reports are welcome as GitHub issues.
See the release notes for what changed in each release.