Don't believe the paper —
recompute it yourself
DDX54, the gene a 2025 PNAS paper names as the commander of immune evasion in lung cancer — you rerun that paper's central experiment from the public raw data, from scratch. Without looking at the authors' numbers, you do the normalisation, the statistics and the pathway analysis with your own hands.
Python runs right inside this page. There is nothing to install.
How you'll learn
RNA is a photocopy of the blueprint, CPM is how many solids there are per 100 mL of broth, GSEA is a team's average score. You get a picture of the meaning before you memorise the formula.
Not a stand-in dataset — the actual raw data from GEO (GSE285342), exactly as deposited. Every piece of code has a run button.
You recompute from scratch without seeing the authors' fold changes or p-values, and reach the verdict yourself on whether the same conclusion comes out.
Every lesson has a laboratory widget — turn the filter thresholds, the shrinkage strength and the volcano-plot cut-offs yourself.