The Ovarian Pre-cancer Atlas is developed by the Shih Laboratory at Johns Hopkins University to share spatial transcriptomic insights on precursor lesions of ovarian cancer. Our goal is to provide the research community with intuitive tools to explore gene expression patterns across the fallopian tube epithelium and associated precancerous lesions.
You may search genes of interest to inquire mRNA expression levels and compare them in normal fallopian tube epithelium, high-grade serous carcinoma, and various types of precancerous lesions including STIC and p53 signatures. In the forthcoming version, the integrated morphology, multi-omic data, and relevant clinical information will be available.
Gene Explore
Gene Expression
Compare a gene’s RNA expression across lesion groups in the epithelium or stroma. Add a morphology comparison to examine Flat and BLAD lesions.
How to use
Explore modeled RNA expression across normal fallopian tube epithelium, precursor lesions, and high-grade serous carcinoma using NanoString GeoMx DSP data.
- Select Type: Choose Epithelium or Stroma to define the tissue compartment.
- Enter Gene: Type an official gene symbol (e.g., SOX4, MMP7, CD74).
- (Optional) Select Category: Add an additional plot to compare STIC separated by morphology category (Flat and BLAD).
- Click Generate Plot: Boxplots will display expression levels across lesion groups.
Note on “*”: The asterisk marks groups whose group mean is above or below the NFT (normal fallopian tube epithelium) mean based on a simple reference comparison computed in Python. It is a visual cue only and does not indicate statistical significance (no p-value). For formal differential expression or significance testing, please use DESeq2 (or an equivalent method).
Gene Explore
Gene Correlation
Examine the relationship between two genes in epithelial RNA data. Choose the groups to include and optionally color samples by diagnosis, molecular subtype, or morphology.
How to use
Explore the relationship between two genes across selected lesion groups using mRNA expression data from NanoString GeoMx spatial transcriptomics.
- Select Category: Choose a grouping variable (e.g., Diagnosis, Morphology category, or Molecular subtype).
- Select Group(s): Choose one or multiple lesion groups to include in the correlation analysis.
- Color By (Optional): Highlight points by Diagnosis or Molecular Subtype.
- Enter Gene A and Gene B: Input two official gene symbols (e.g., IGFBP2 and MMP7).
- Click “Generate Correlation”: A scatter plot will display gene expression correlation across the selected groups.
Summary
Correlation summary
COPY NUMBER & RNA
Copy number & gene expression
Cohort & data availability
COHORT OVERVIEW
Copy-number profiles with representative RNA annotations
Browse the 48 currently included DNA–RNA links, chromosome-arm Z-scores, and their representative RNA regions.
Three outliers in the NFT group have been removed.
REPRESENTATIVE RNA AVAILABILITY
Included representative RNA regions by diagnosis
RNA totals count regions. Included representative regions are selected one per DNA sample; multiple regions may belong to one case.
| Diagnosis | RNA regions | Included representative RNA regions | % of RNA regions |
|---|---|---|---|
| Loading availability… | |||
RNA analysis controls
ALIGNED SAMPLE VIEW
Chromosome-arm aneuploidy, gene copy number & RNA expression
Arm Z-scores use a continuous diverging color scale centered at zero and capped visually at ±20; the paper’s aneuploidy threshold is |Z| > 5. Gene-locus DNA bars show supplied segment Log Ratios on a separate symmetric scale with display guides at ±0.25. RNA points show processed expression values; stems connect each point to the NFT median. The shaded band describes the middle 50% of 99 NFT epithelial regions.
GENE POSITION → DNA SEGMENTS
Copy-number signal at gene locus
Choose a gene to inspect its genomic position.
COHORT CONTEXT
Expression distribution
Bar height shows relative RNA-region count; representative observations are included in the reference. This is a descriptive view.
Data & methods
INTERPRETATION & PROVENANCE
Data & methods
Measurements
Chromosome-arm values are provided Z-scores; colors are capped visually at ±20 and the paper’s aneuploidy threshold is |Z| > 5. Segment profiles show supplied Log Ratios at hg19 coordinates. The ratio base and endpoint conventions are not documented in the supplied source files; no fold-change or absolute-copy conversion is applied. Unresolved boundaries and genes without covered segments are not assigned zero. Segment display guides are ±0.25.
Representative RNA region selection
The original analysis selects one representative RNA region per DNA sample by ordering first on molecular subtype (NFT, Dormant, Mixed, Immunoreactive, Proliferative, HGSC), then on Diagnosis (NFT, p53, STIL, STIC, HGSC; other values last), and keeping the first. The displayed subtype and diagnosis are annotations from that selected region; STIC-like diagnoses are displayed as STIC.
Linkage and availability
The source snapshot contains 51 DNA samples and representative RNA regions across 29 cases of the 68-case RNA cohort. This view includes 48 links across 28 cases after removing three NFT outliers. FTE labels use the sample-name matches in the source metadata table. One included sample (FTE-038 · CNV-049) is labelled by case and sample key because its lesion identifier is unavailable. DNA and RNA were obtained from adjacent tissue sections. The RNA reference distributions cover 294 regions, including 99 NFT epithelial regions; selected RNA values are included in those distributions. The study’s sample-level expression matrix was variance-stabilized and batch-corrected with limma; this view applies no additional transformation.
Scope
This descriptive view shows relative copy-number signals and RNA expression. Mutation status and absolute copy counts are not shown. It does not calculate prediction, RNA correlation, or an aneuploidy burden index and is not for clinical decisions.
Study context: Chang et al., Integrated Spatial Analysis Reveals the Molecular Landscape of Ovarian Precancerous Lesions.