There is a familiar problem in data science: what do you do when there is nothing directly comparable—no reference population, no matching case and no benchmark?

In rare and ultra-rare disease, the absence of a comparator can quietly turn the word rare into a substitute for explanation. But “rare” is not a diagnosis. It describes the scarcity of reference data; it does not end the duty to examine partial matches, mechanisms, intervals and the full clinical history.

Partial evidence can still be useful

A rigorous search does not need an identical case in order to move forward. It can compare symptoms, chromosomal regions, inheritance mechanisms and biological plausibility—provided every connection is labelled as a hypothesis rather than a conclusion.

Correction is part of the method

An early hypothesis in S.F.'s case focused on RFX3 because the published phenotype appeared strikingly similar. Later coordinate review demonstrated that RFX3 lies outside the corrected deletion interval. The hypothesis was withdrawn.

The aim was never to prove the first idea right. It was to get as close as possible to what the data could honestly support.