Overview

scSpAS is an event-centered atlas for alternative splicing (AS), alternative polyadenylation (APA) and alternative transcription start sites (ATS) in single-cell and spatial datasets.

Start from the Home search box or Browse page using a gene, event ID or genomic coordinate. Select an event row, then open its Cell, Spatial or Regulation view. The event context is retained when moving between compatible pages.

Search & Browse

The Home search and Browse search accept gene symbols/IDs, native event IDs, genomic coordinates and indexed biological terms. The expandable Batch search accepts up to 500 genes, event IDs or coordinates.

1. Choose filters such as species, tissue, condition, data modality, technology and event type. Advanced filters contain method, dataset, sample and genome assembly.
2. Click Apply Filters. Available choices update according to the current combination.
3. Click a sortable column heading to reorder results, or use Previous, Next and Rows per page to navigate.
4. Use the Cell, Spatial or Regulation button in the Open column to examine that exact event observation.

A disabled Open button means that the corresponding evidence layer is unavailable for that dataset observation. Fresh Browse sessions initially show coordinate-style events; use search or filters to refine the result set.

Cell View

Select a Dataset, identification Method, cell Annotation and Event, then click Apply. The structure panel opens around the selected event; use Event, Full gene, pan and zoom controls to explore the complete gene region. The page also displays annotation UMAP, event usage, host-gene expression and group-level boxplots. Isoform panels appear only when isoform data are available.

Cell-type-specific events

Expand Cell-type-specific events to inspect whether the selected event is enriched or depleted in a cell type and to browse the top significant events. The table reports the cell type, Δ usage and BH-adjusted FDR; click an event row to open it, or use Download all results to export the complete result set.

Cell type-level pseudobulk

When matching pseudobulk matrices are available, Cell Type-level Pseudobulk Analysis aggregates the selected event and its host-gene expression by the chosen annotation. It provides a cell-type-colored UMAP, event usage by cell type and host-gene expression by cell type. The section is hidden when the selected dataset or method has no compatible pseudobulk data.

ATS RNA degradation and integrity

For scATS datasets with degradation/integrity metadata, the Event Degradation–Integrity Landscape is displayed automatically. The x-axis shows RNA degradation rate (α), the y-axis shows transcript integrity (β), and the selected ATS event is highlighted against the dataset-wide distribution. The marginal density plots summarize α and β separately. If this panel is absent, compatible α/β source data are unavailable; the values should not be interpreted as zero.

In correlation sections, click Calculate to rank genes or events by Spearman correlation. Export downloads the full ranking. Selecting a target gene/event updates its UMAP and scatterplot.

The GSEA barplot uses the correlation-ranked gene list and WikiPathways: the x-axis is NES, and bar color can be switched between −log10(q value) and −log10(p value).

Spatial View

Select a spatial Dataset, identification Method, grouping Annotation and Event, then click Apply. The structure panel opens around the selected event; use Event, Full gene, pan and zoom controls to explore the complete gene region. Region annotation is used by default when available.

Initial loading

The three main maps show annotation, event usage and host-gene expression on matched tissue coordinates. Regional boxplots summarize the same values by the selected annotation; optional isoform maps appear when available.

Annotation-specific events

Expand Cell-type-specific events to evaluate events across the currently selected spatial annotation, such as a region, layer, cluster or cell type. The selected-event summary and top-events table use the same one-versus-rest Δ usage, sample-size and BH FDR rules described for Cell View. Positive Δ usage means higher event usage in the named annotation group; click a table row to open that event or download the full result set.

Region-level pseudobulk

When compatible pseudobulk data are available, Region-level Pseudobulk Analysis aggregates event usage and host-gene expression by the selected spatial annotation. It shows the aggregated value on the tissue coordinates and compares event usage and host-gene expression across regions. If the active grouping field contains cell types, the heading changes to Cell Type-level Pseudobulk Analysis. The section is hidden when no matched pseudobulk matrix is available.

Event–gene correlation, GSEA, target-gene selection, event–event correlation and export work in the same way as in Cell View, but use matched spatial spots.

Regulation

Select the dataset, method, annotation and event, then click Apply. All tracks are aligned to the same genomic ruler and automatically match the event species and assembly.

Click an evidence button to show or hide its track; expandable buttons allow individual tracks to be selected. Use arrows to pan, +/− to zoom, Reset to restore the locus and Track order to rearrange active groups. Hover over features for labels and coordinates.

CLIP peaks additionally report RBP/sample metadata and log10 p.value. The candidate RBP table can be sorted by its numeric columns.

Download

Filter files by data modality, technology, species, condition, event type, dataset or file category, then click Apply. Reset clears all filters. Dataset cards summarize availability and provide direct file downloads.

Standard categories are event matrices, normalized gene-expression matrices, isoform matrices and cell/spot metadata. Not every dataset contains every category.

Notes

If a search returns no records, clear filters and verify the method-specific event ID. For data problems, include the dataset, method, complete event ID and genome assembly when contacting the team.