How we work

1

Consultation & experimental design

Every collaboration starts with a conversation, not with data. We discuss your biological question, what you already know, and what an answer would look like. From there we work backwards: which measurements can actually answer the question, how many samples are needed for statistical power, which confounders should be controlled at the design stage.

The most expensive mistakes in omics are made before any data is generated, and they are the hardest ones to fix afterwards.

2

Data quality control

Before any analysis, we examine what the data actually contains: read quality, contamination, batch effects, outliers. You receive a QC report that states clearly what the data can support and, just as importantly, what it cannot.

If something is wrong, this is where it surfaces, while there is still time to act on it. No result is built on data we have not inspected first.

3

Analysis & statistics

The analysis itself is built around your question, not adapted from a template. We select and test the methods against the nature of your data, and when the standard approach is not the right one, we say so and explain why.

Pipelines are versioned and reproducible: every parameter, threshold and software version is documented. Anyone – you, a colleague, a reviewer – can trace how each number was produced.

4

Biological interpretation

Tables and plots are not answers. We meet with you to discuss what the results mean, place them in the context of your field’s literature, and separate solid signals from artefacts.

This stage is a dialogue: your knowledge of the system and our knowledge of the data, until the biological story holds.