Single-Cell RNA-Seq Data Analysis Service
CD Genomics offers a comprehensive Single-Cell RNA-Seq Data Analysis Service that transforms raw scRNA-seq or snRNA-seq data into publication-ready biological insights. Unlike automated cloud platforms that require your team to make analysis decisions, our expert bioinformaticians design and execute every step of the pipeline — from QC through advanced trajectory and cell–cell communication analysis — delivering interpreted results, not just output files. The service is platform-agnostic, supporting data from 10x Genomics Chromium, BD Rhapsody, Smart-seq2/3, Drop-seq, Parse Biosciences, and other single-cell technologies.
- Platform-agnostic: supports 10x, BD Rhapsody, Smart-seq, Drop-seq, Parse, and more
- Modular pipeline: choose standard analysis, advanced modules, or a fully customized workflow
- Expert-driven, not automated: experienced bioinformaticians design your analysis, not a black-box pipeline
- Publication-ready output: fully documented code, methods write-up, and journal-quality figures
Services are provided for research use only.
Standard Analysis Pipeline
Our standard pipeline covers every step from raw data to annotated cell clusters and differential expression. Each module includes QC checkpoints and produces publication-ready figures with fully documented, reproducible analysis code (R/Seurat or Python/Scanpy at your preference).
Preprocessing & QC
Raw FASTQ files are processed through platform-specific pipelines — Cell Ranger (10x), BD Rhapsody Pipeline, STARsolo (Smart-seq/Drop-seq), or kallisto bustools. Cell-level QC includes gene counts, molecular tag counts, mitochondrial/ribosomal fractions. Doublets and low-quality cells are filtered using adaptive thresholds with full documentation.
Normalization & Integration
SCTransform (Seurat v5) or LogNormalize with Harmony or RPCA batch correction for multi-sample studies. Integration quality assessed via LISI scores and cluster mixing metrics.
Clustering & Annotation
PCA, UMAP, t-SNE embeddings. Louvain/Leiden clustering at multiple resolutions with stability assessment. SingleR + CellTypist automated annotation with manual curation using canonical markers. Cell-type proportions quantified and compared across conditions.
Differential Expression
Wilcoxon rank-sum (presto) or pseudobulk DESeq2 for multi-sample designs. Volcano plots, heatmaps (ComplexHeatmap), dot plots. Markers ranked by log fold change and AUC.
Gene Set Enrichment
clusterProfiler against GO, KEGG, Reactome, Hallmark, WikiPathways. Dot plots, enrichment maps, ridge plots with FDR-adjusted p-values.
Reproducible Output
Interactive HTML reports, fully documented R/Python scripts with session info, enabling independent reproduction of all analyses.
Advanced Analysis Pipeline
Our advanced modules go beyond cell-type classification to extract mechanistic insight from your single-cell data. Each module is independent — you choose only what your research question requires.
Trajectory & Pseudotime
Monocle3 or Slingshot reconstruction of developmental, differentiation, and activation trajectories. Gene expression heatmaps along pseudotime, branch expression analysis modeling (BEAM), identification of dynamic and branch-point genes.
RNA Velocity
scVelo with dynamical modeling using spliced/unspliced counts from velociraptor/velocyto. Velocity vectors on UMAP reveal transitioning vs. steady-state populations — a temporal dimension invisible to static clustering.
Cell–Cell Communication
CellChat v2 ligand–receptor analysis: circle plots, chord diagrams, pathway-level signaling summaries, pattern recognition across conditions, and spatial-aware communication modeling.
TF Regulatory Networks
pySCENIC transcription factor activity and regulon inference. AUCell scores quantify regulon activity per cell. Identifies which TFs drive specific cell states and how regulons vary across conditions.
Tumor CNV & Multi-Sample
inferCNV for chromosomal copy number alterations, clonal substructure, malignant vs. non-malignant classification. Pseudobulk frameworks for statistically robust cross-condition comparisons.
Custom Visualization
Publication-quality multi-panel figures, graphical abstracts, spatial-aware layouts, and interactive HTML dashboards tailored to your manuscript’s narrative.
Data Deliverables
Every project includes structured, ready-to-use deliverables.
- Processed data
- Filtered and normalized gene–cell expression matrices (.h5ad, .rds, .h5), cell metadata with annotations and cluster assignments
- QC report
- Per-sample and per-cell QC metrics with visual summaries, doublet detection report, integration quality assessment
- Analysis report
- Complete results for all selected modules — clustering, annotation with marker validation, differential expression, enrichment, and all advanced analyses selected
- Publication-ready figures
- Vector (PDF/SVG) and raster (PNG/TIFF at 300+ dpi) formats for every analysis output, with editable source files
- Reproducible analysis code
- Fully documented R/Python scripts with session information (sessionInfo/conda environment), enabling independent reproduction of all analyses
- Methods write-up
- A draft methods section describing all analytical steps, parameters, and software versions — suitable for direct inclusion in your manuscript
scRNA-seq Data Analysis Applications
Our scRNA-seq data analysis service supports a wide range of biological research questions.
Cell atlas construction
Build comprehensive cell-type maps by integrating multiple scRNA-seq datasets with robust batch correction and consistent annotation. Our team has experience with atlas-scale projects spanning dozens to hundreds of samples.
Disease mechanism studies
Identify disease-associated cell states, perturbed pathways, and dysregulated ligand–receptor interactions by comparing patient and control samples at single-cell resolution.
Developmental and differentiation trajectories
Reconstruct lineage relationships, identify transitional cell states, and discover fate-determining transcription factors during development or drug-induced differentiation.
Tumor microenvironment characterization
Map tumor-infiltrating immune and stromal populations, infer copy number alterations in cancer cells, and decode immunosuppressive signaling networks. See also: Tumor Microenvironment Solutions.
Drug response and pharmacodynamics
Compare scRNA-seq profiles across treatment conditions to identify drug-responsive cell populations and resistance-associated transcriptional programs.
Multi-omics integration
Combine scRNA-seq with spatial transcriptomics, scATAC-seq, or bulk RNA-seq for integrated multi-dimensional analysis. See also: BD Rhapsody scWTA Service for experimental services.
Case Study: Multi-Omics Integration Reveals Epithelial Heterogeneity in Bladder Cancer
Source: Integrated single-cell and spatial transcriptomics combined with whole-exome sequencing reveal key hub genes and epithelial heterogeneity in bladder cancer (Li L, Li Q, et al., Frontiers in Oncology, 2025)
Background
Bladder cancer exhibits substantial tumor heterogeneity driven by distinct epithelial subpopulations with varying degrees of stemness and differentiation potential. Identifying the specific subpopulations linked to tumor mutation burden (TMB) and poor prognosis requires simultaneous analysis of multiple data modalities and advanced computational methods beyond standard clustering.
Methods
The study analyzed three integrated data layers: single-cell RNA-seq (13 samples, 77,263 cells), spatial transcriptomics, and whole-exome sequencing (30 samples) alongside 514 bulk transcriptomes. The bioinformatics pipeline included UMAP-based clustering with secondary subclustering, CellChat for intercellular communication network inference, CytoTRACE for stemness scoring, and Monocle2 pseudotime trajectory analysis to reconstruct epithelial differentiation paths. TMB was linked to cell subsets via scAB and Ro/e algorithms. A random survival forest model identified prognostic hub genes, experimentally validated by qPCR and Western blot.
Results
A specific epithelial subpopulation (Epi14) was strongly associated with elevated TMB and poor clinical outcome. Pseudotime trajectory analysis revealed that Epi14 sits at the origin of a differentiation path toward more mature epithelial states. Cell–cell communication mapping identified the signaling networks through which Epi14 interacts with the broader tumor microenvironment. The machine learning-driven analysis yielded a validated risk-score model and identified ABRACL and ARPC3 as clinically relevant hub genes.
Conclusion
This study illustrates how comprehensive scRNA-seq data analysis — spanning clustering, pseudotime reconstruction, cell–cell communication, stemness scoring, and machine learning-based feature selection — can transform raw sequencing data into translational insights. The multi-omics integration and analytical depth demonstrated in this work mirror the end-to-end bioinformatics service CD Genomics provides for single-cell studies.
Frequently asked questions (FAQ)
References
- Hao Y, et al. "Integrated analysis of multimodal single-cell data." Cell, vol. 184, no. 13, 2021, pp. 3573–3587.e29.
- Wolf FA, Angerer P, Theis FJ. "SCANPY: large-scale single-cell gene expression data analysis." Genome Biology, vol. 19, 2018, 15.
- Bergen V, Lange M, Peidli S, Wolf FA, Theis FJ. "Generalizing RNA velocity to transient cell states through dynamical modeling." Nature Biotechnology, vol. 38, 2020, pp. 1408–1414.
- Li L, Li Q, Liu J, Wang Y, Ma Y, Tang L, Zhao Y. "Integrated single-cell and spatial transcriptomics combined with whole-exome sequencing reveal key hub genes and epithelial heterogeneity in bladder cancer." Frontiers in Oncology, vol. 15, 2025, 1748105.
