How many samples do I actually need for RNA-Seq?
Power depends on effect size, biological variability and sequencing depth — not on a rule of thumb. A practical walkthrough with worked numbers for common designs.
Resources & Guides
- Home
- Knowledge Hub
- Resources
Practical guides on the analysis decisions that most often decide whether a study works.
These guides are being written now and will be published here as they are finished. The topics below are the questions that come up in almost every scoping call. Everything will be free, and none of it will require you to become a client.
If a guide you need is missing, tell us — we write them in response to what people actually ask.
Guides and Explainers
Written by the consultants who do the work.
Batch effects: how to design them out before you sequence
The single most common reason an expensive experiment yields nothing interpretable, and the randomisation strategies that prevent it.
Why variant pipelines built on European data fail on African samples
Reference bias, allele frequency filtering and annotation gaps — what goes wrong, and the concrete changes that fix it.
A reproducible analysis checklist for research groups
Version pinning, environment capture, seeds and file manifests. What to do so your result survives a reviewer asking you to rerun it.
Are your clusters real? A practical QC checklist
Ambient RNA, doublets and resolution choice routinely produce clusters that do not exist. How to tell the difference before you annotate them.
Compositional data: why standard tests give you the wrong answer
Microbiome abundances are proportions, and proportions break the assumptions of the tests most commonly applied to them. What to use instead.
Reading a survival curve properly
Censoring, competing risks and the proportional hazards assumption — the three things reviewers check and authors most often miss.
Your model beats the baseline. Does it?
Data leakage, optimistic validation and the subgroup performance question. A short guide to not fooling yourself.
Naming files so your pipeline does not break
Spaces, ampersands and inconsistent sample IDs cause more failed runs than any algorithm. A conventions guide you can hand to your lab.
Reference Material
Templates and checklists
Documents you can download and adapt for your own group:
- Sample sheet template for sequencing submissions (CSV)
- Statistical analysis plan template for clinical studies
- Data management plan template for grant applications
- Pre-submission analysis checklist
- Reproducibility checklist for manuscripts
External resources we recommend
Not ours, but consistently good:
- H3ABioNet — training materials and node network across Africa
- Bioconductor and Galaxy training — free, well-maintained tutorials
- The Carpentries — foundational data and software skills workshops
- nf-core — community-curated, peer-reviewed Nextflow pipelines
- Nextstrain — real-time pathogen evolution tracking
Ask us directly
If you have a question that a guide would not answer, send it to our team. We answer research questions from African students and researchers whether or not there is a project attached.
