Written for people who have the data and have to make decisions about it. No sign-up, nothing gated. RSS.
Power thinking for single-cell studies: how many cells per sample, how many reads per cell, and why the number of subjects matters more than either.
What each spatial technology actually measures, what resolution really means in practice, and which question each one can and cannot answer.
What to put in the budget for analysis, how to justify it to a review panel, and the line items applicants forget until the money is already allocated.
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.
How to tell a real rare population from a doublet cluster or ambient contamination, which tools to use, and the order to run them in.
How to annotate single-cell clusters so the labels survive peer review — the failure modes of each approach and a workflow that combines them.
How to choose an integration method, how to tell over-correction from correction, and the checks that show your biology survived.
Why per-cell differential expression inflates false positives, when pseudobulk is the right answer, and exactly how to run it.
Monocle3, Slingshot and RNA velocity produce a trajectory from almost any dataset. How to tell whether yours reflects biology.
Ligand–receptor scoring is a hypothesis generator, not evidence of interaction. How to run it defensibly and how to report it.
What deconvolution, spatially variable genes and niche analysis actually do, which platform needs which method, and where spatial analyses usually go wrong.
A practical route from a GEO accession to an analysis you can trust — what to check in the metadata, when raw counts matter, and the traps in combining datasets.
A file-by-file walkthrough of the 10x Cell Ranger output directory — what each one is for, which to keep, and which to send when someone asks for your data.
A practical comparison of the two dominant single-cell toolkits — where each is genuinely better, where results differ, and what that means if you switch mid-project.
An honest comparison of the four ways to get single-cell data analysed — including when each one is the wrong choice.
Twelve questions that separate a deliverable you can publish from a folder of plots — data handling, reproducibility, revisions, and who owns what.
The file formats, metadata and written questions that turn a two-week analysis into a three-day one — with a copy-paste checklist.
Ten checks you can run in an afternoon — without being a bioinformatician — to find out whether a delivered single-cell analysis will survive review.
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.
Timelines for each stage of a single-cell project, what actually causes delays, and how to compress a deadline without cutting corners.
How to look for a 0.5% population in single-cell data, and the four artefacts that produce a convincing rare cluster that is not real.
Consent, controlled access, transfer mechanics and deletion — the practical checklist for getting human sequencing data to someone outside your institution without creating a problem.
What reviewers check in a single-cell methods section, a fill-in template covering every parameter that gets asked about, and the omissions that trigger revision.
Chromatin accessibility data has its own quality metrics and its own normalisation, and one routine step silently changes the cluster boundaries. What to check, with a measurement on public data.
Cell counts do not substitute for biological replicates. What the number of donors actually buys you, why pooling cells inflates significance, and how to plan a design that survives review.
Spatial clustering always returns domains, whether or not they mean anything. Four checks that separate a real niche from a quality gradient, with a worked example where one of our own domains failed.
Deconvolution is far cheaper than new single-cell experiments and is genuinely sufficient for some questions. A benchmark with known ground truth showing exactly where it holds and where it breaks.
Antibody counts are not expression counts. The isotype-control number, the normalisation, and the protein-versus-RNA check that decides whether a surface marker means anything — with numbers from a public dataset.
Converting between Seurat and h5ad, what silently goes missing on the way, and the two-line fallback that works when nothing installs.