Analysis-only bioinformatics for single-cell, spatial and bulk transcriptomics. No sequencing contract, no minimum order — send a count matrix or an accession number and get back publication-ready results, the code that produced them, and a methods paragraph. Fixed price agreed before we start.
A reference classifier run without seeing our manual labels agreed on 6/6 clusters, 100% within each. See it
Pseudobulk vs per-cell on 8 donors: the two gene lists disagree on 68.3% of calls. See it
Donor clustering down 20.2%, treatment effect moved -0.1%. Measured, not asserted. See it
Open a full example report — the actual deliverable, figures embedded, methods paragraph included. Every number above was computed by a script you can download and re-run (the code is here). We publish the checks other vendors leave out — including when they are unflattering.
Five complete sample deliverables, all run on public data and fully reproducible: single-cell (2,638 cells, 6 annotated populations) · spatial (4,025 Visium spots, 10 spatial domains) · bulk RNA-seq (DESeq2, 2,414 genes up, with a biological check that passes) · pseudobulk vs per-cell (8 donors; the two methods disagree on 68.3% of calls) · differential abundance (pooling cells calls 4 cell types as changed; a donor-level test calls none) · trajectory (PAGA + pseudotime, with the branch check that most reports skip).
QC, integration, clustering, annotation, differential expression, trajectory, cell–cell communication. 10x, BD, Parse, Singleron, plate-based.
Visium / Visium HD, Xenium, Stereo-seq, MERFISH: spatially variable genes, deconvolution, niche analysis, tissue overlays.
DEG, GO/KEGG/GSEA, WGCNA, survival association; ATAC, CUT&Tag and multi-omics integration on request.
Got a "please re-analyse" from reviewers? We reproduce, correct and extend an existing analysis, with a written response to each comment.
Every project ships containerised code and parameters. Your students can re-run it after we are gone — that is the point.
NDA on request, data deleted after delivery, nothing published or reused. Unpublished data stays unpublished.
Upload a count matrix and get back barcode-rank and per-cell distributions, thresholds computed from your data rather than a tutorial, the genes that dominate your counts, an expected doublet count, and a downloadable per-cell table. No account, no email, and the file is deleted when the report is built. Run it →
Practical write-ups from the same people who would do your analysis — read them before you decide anything.
A practical triage for the five re-analysis requests reviewers actually make, what each one really costs you in time, and how to answer without redoing the whole project.
Honest price ranges for scRNA-seq analysis in 2026 — in-house, core facility, freelancer and service provider — and where the hidden costs are.
Why fixed cut-offs like "<5% mitochondrial" fail on some tissues, how to set data-driven thresholds, and what to report so reviewers stop asking.
Indicative ranges in USD. You get a fixed written quote before any work starts — no hourly billing, no surprises.
QC, filtering, normalisation, HVG, PCA/UMAP, clustering, marker genes, automated + manual cell-type annotation, publication-ready figures, methods text.
Integration/batch correction across samples, differential abundance, pseudotime/trajectory, cell–cell communication, gene-set/pathway scoring, doublet and ambient-RNA handling.
Spot/cell QC, clustering, spatially variable genes, deconvolution against a reference, niche/neighbourhood analysis, overlay figures.
Alignment or count-matrix intake, QC, normalisation, differential expression, GO/KEGG/GSEA enrichment, volcano/heatmap figures, results tables.
Independent re-run of an existing dataset, reproduction of published figures, targeted answers to reviewer comments, revised figures.
Bespoke workflows, multi-omics integration, reproducible Nextflow/Snakemake delivery, containerised handover so your team can re-run everything.
Working in industry rather than academia? What companies ask before price — NDA, auditability, fixed dates, pilot first.
Answer four questions and get a concrete plan — the modules that fit your design, what to send, and an indicative price. It will also tell you if your design cannot support the comparison you are planning, which is worth knowing before you pay anyone. Build a plan →
New to working with an external analyst? Here is exactly how a project runs — including what happens if your data fails QC, what we do with it afterwards, and who owns the results.
Tell us what you have. You get a scoped fixed price and a delivery date, normally within one working day.
We will come back with a fixed scope, price and delivery date, normally within one working day.