Technology & platform

Scientific capabilities

Single-cell & single-nucleus RNA-seq

Quality control, normalization, dimensionality reduction, clustering, cell-type annotation and differential expression. Trajectory inference, RNA velocity and CellRank support analysis of cell fate, alongside regulatory network and cell–cell communication analysis.

Spatial transcriptomics

Spatial data processing, domain identification, deconvolution, spatial autocorrelation, spatially variable gene detection and spatial cell–cell communication. Cell-type expression profiles are used to estimate abundance proportions within spatial units and map single-cell subtypes to spatial coordinates.

Cross-modality integration

Integrating the transcriptomic depth of single-cell and single-nucleus data with spatial context through label transfer, reference-based deconvolution and cross-platform validation.

Cell–cell communication

Multi-method aggregation of ligand–receptor interaction scores across cell-type pairs. In spatial data, proximity weighting is used to assess whether inferred signaling is supported by neighboring cell locations.

Regulatory network analysis

Inference of transcription factor activity and gene regulatory networks, projected into spatial coordinates to connect regulatory activity with tissue context. Virtual gene knockout is used to predict downstream effects of perturbing specific regulators.

Computational approaches

From reference construction to integrated interpretation, the analysis is shaped by the research question and the data available.

Reference construction & annotation

Standardization, dimensionality reduction and graph-based clustering identify cell populations. Marker-based annotation is interpreted in the relevant tissue and disease context.

Spatial validation & consistency

Spatial autocorrelation assessment, comparison of temporal trends between platforms and visualization of ligand–receptor co-expression in spatial coordinates.

Public data & cross-dataset integration

Retrieval and analysis of public data from GEO, CNSA and STOmics DB. Integration across platforms and batches is preceded by assessment of model consistency, time-point overlap, tissue-region compatibility and data scale.

Computational strategy & analysis workflows

Analysis strategy is designed around the biological question and dataset compatibility. Workflows include reusable code, parameter documentation and reproducibility materials for research collaboration.

Tools & environments

Examples of the computational environment; selection depends on the analysis task.

R ecosystem

Seurat, RCTD, Spacexr and related packages.

Python ecosystem

Scanpy, Squidpy, scVI, scVelo and related frameworks.

Communication analysis

LIANA+ and multi-method ligand–receptor inference.

Spatial statistics

Moran’s I, spatial weighting and neighborhood-based metrics.

Representative analytical outputs

The following describes output types within these capabilities. No project results or example datasets are displayed here.

Cell-type & subtype maps

Marker-supported annotations, differential expression, GO enrichment and GSEA, with multiple-testing correction and effect-size metrics.

RNA velocity & cell fate analysis

Trajectory inference and CellRank-based cell fate probability estimates, with cross-validation and consistency checks across methods.

Spatial abundance matrices

Bin- or spot-resolution deconvolution outputs with uncertainty estimates and goodness-of-fit diagnostics.

Spatially weighted communication networks

Ligand–receptor interaction scores, spatial co-localization significance and cross-platform consistency metrics.

Spatial regulatory activity maps

Transcription factor activity and regulatory networks in spatial coordinates, with statistical support for activity and spatial enrichment.

Figures & documented data resources

Analytical figures, code, parameter documentation and reproducibility materials, including documentation of batch-effect handling.