CLAIREscope: Cellular Landscape Analysis, Interpretation & Results Explorer

Abstract

CLAIREscope (Cellular Landscape Analysis, Interpretation & Results Explorer) is an open-source interactive platform for reproducible single-cell transcriptomics analysis, built with Python, Scanpy, and Streamlit. It integrates data preprocessing, multi-sample exploration, statistical analysis, trajectory modeling, differential expression, pathway enrichment, and publication-ready output generation within a unified analytical workflow. The current release implements 11 specialized interactive analysis studios: Static UMAP Explorer — fixed-aspect UMAP visualization, continuous gene-expression mapping with percentile-based clipping, multi-sample split views, and coordinate/data export. Interactive UMAP Studio — Plotly-based cell-level exploration with dynamic cohort filtering, population dimming, and hover-level inspection of expression and metadata. Sample Composition & Stratification — calculation and visualization of absolute cell counts and relative cell-state frequencies across samples and experimental conditions. Gene Expression & Statistical Testing — violin-based expression analysis with user-defined group comparisons, Mann–Whitney U/Wilcoxon rank-sum testing, automated significance annotation, and optional zero-expression filtering. Gene Signature & Pathway Scoring — dynamic Scanpy-based scoring of built-in or user-defined gene sets, with cross-condition statistical comparison within selected cell populations. Co-expression & Correlation Analysis — gene–gene, gene–signature, and signature–signature analysis with Pearson and Spearman correlations, regression trends, and configurable dropout filtering. Continuous Trajectory & Spline Modeling — modeling of gene-expression and signature dynamics along diffusion pseudotime (DPT) using polynomial regression and B-spline fitting. Differential Expression & Volcano Analysis — bidirectional Wilcoxon differential expression analysis with multiple-testing correction and interactive, user-adjustable volcano-plot thresholds and gene labeling. Clustered Heatmap Analysis — hierarchical clustering using pairwise distance and linkage methods, Z-score standardization, and interactive sample/cell-state ordering. Pathway Over-Representation Analysis (ORA) — hypergeometric enrichment testing against Gene Ontology Biological Process, KEGG, and Reactome, including separate analysis of up- and down-regulated gene sets. Bulk Packaging & Provenance — batch generation of publication-ready SVG/PDF vector graphics, 300-DPI raster figures, and structured statistical CSV matrices, together with reusable configuration import for reproducible analytical workflows. Rather than functioning solely as a visualization interface, CLAIREscope is designed to support the iterative analytical cycle of exploration → quantitative testing → parameter refinement → biological interpretation → publication-ready export. Analytical parameters, cohort definitions, comparisons, and visualization settings can be adjusted interactively with immediate recalculation, allowing exploratory decisions to be progressively refined into standardized outputs suitable for downstream analysis, reporting, and manuscript preparation.