A gene with zero papers.
How did it rank first?
Idiopathic pulmonary fibrosis has 4,591 candidate targets. Rank them by evidence and telomere genes rise to the top β and the ten seats below them are filled by targets that already have a drug.
TNIK, the one that actually became a drug, is not on that list at all. PubMed counts zero IPF papers for it at the time of discovery.
You rebuild the ranking in Python against three live APIs, then go to check causality with Perturb-seq β and meet the wrong cell.
How you'll learn
Open Targets, PubMed, Perturb-seq. All three wrappers keep the real contract.
You count TNIK's IPF papers on PubMed. At discovery time the answer is zero.
Same data, same code, different winner. What decided it was not the arithmetic.
You knock TNIK down and 11 genes move. The experiment was fine β the cell was wrong.