Experimental Design

From biological question to experiment

Strong experimental design is the difference between data that merely exists and data that can confidently answer a biological question. I work with you from the earliest planning stages to ensure your experiment is statistically robust, biologically meaningful, and aligned with downstream analysis and publication or funding requirements.

This service is particularly valuable before samples are collected, but it can also be applied to refine ongoing projects or to rescue underpowered or poorly controlled designs.

What I help you design

1. Clear biological questions & hypotheses

We start by translating broad aims into explicit, testable hypotheses. This step shapes every downstream decision: assay choice, controls, replication strategy, and analysis plan.

2. Assay selection & strategy

Guidance on choosing and combining genomics assays, including:

RNA-seq (bulk and time‑course)
ChIP-seq / CUT&RUN / CUT&TAG (TFs, histone marks, spike‑ins)
ATAC-seq
qPRO‑seq / nascent transcription assays

Where appropriate, I help design multi‑omic strategies that maximise information while controlling cost and complexity.

3. Replication, controls & confounders

I provide explicit recommendations for:

Biological vs technical replicates
Positive and negative controls
Batch structure and randomisation
Strategies to mitigate confounders (cell passage, donor effects, sequencing runs)

Designs are optimised for statistical power, not just minimum compliance.

4. Power considerations & sample sizing

Where feasible, I incorporate empirical or literature‑based variance estimates to guide:

Minimum sample numbers
Trade‑offs between depth and replication
Risks associated with underpowered designs

The goal is to avoid experiments that are expensive but inconclusive.

5. Analysis‑aware experimental design

Because I also build the downstream pipelines, I design experiments with analysis in mind:

Factorial designs compatible with DESeq2 / edgeR / limma
ChIP‑seq designs suitable for reproducible peak calling and QC
Time‑course and perturbation designs that support interpretable modelling

This avoids common mismatches between experimental structure and analytical assumptions.

Deliverables

Depending on your needs, you will receive:

A written experimental design summary (grant‑ or methods‑ready)
A schematic of the experimental structure
A sample and metadata schema compatible with analysis pipelines
Risk flags and mitigation strategies
Optional alignment with a Snakemake workflow

Who this is for

Academic labs preparing grants or new projects
Industry or biotech teams planning genomics experiments
Researchers transitioning assays into production pipelines
Groups seeking an external, analysis‑literate design review

Why work with me

With over 14 years’ experience spanning wet‑lab research and bioinformatics, I bridge the gap between experimental biology and computational analysis. My designs are:

Practical – grounded in real laboratory constraints
Transparent – assumptions and trade‑offs made explicit
Reproducible – aligned with modern, auditable workflows

If you want your experiment to stand up to reviewers, collaborators, and future you, this is where to start.