Lydon, April 12, 2008-a XBG patient and his parents sued the department of clinical diagnosis in Lydon, the XBG mutation database, and the journal Human Mutation. The complaint was that serious and culpable mistakes were made during the clinical diagnosis of the pregnancy in the XBGfamily, that ultimately led to the birth of an affected child. A paper published in Human Mutation listed the sequence variant detected in the family as "nonpathogenic." Careful examination would have revealed that the change was clearly pathogenic (a nonsense mutation). However, the accused parties failed to verify the data of the original report and just copied it.
Is this imaginary news item pure fiction? Or, might it come true? When a clinical diagnosis is based on the detection of a variant in the DNA sequence (mutation), one wants to be absolutely sure. One of the most reliable decision tools available is to search the literature for confirmative reports. Nowadays, general or gene/diseasespecific databases are often available that make this task rather simple. Consequently, when a sequence variant is detected, it is becoming general practice to check these repositories for previous reports of the change and accept the conclusion submitted by the author, ''pathogenic'' or ''not pathogenic.'' For this process to be reliable, it is critical that mutation reports do not contain errors and that descriptions are unique and unequivocal. For this latter purpose, the HUGO Mutation Database Initiative (MDI) instigated an ad-hoc committee to formulate rules for the description of sequence variants [Beutler, 1993;Beaudet and Tsui, 1993;Beutler et al., 1996]. Based on initial suggestions, the nomenclature committee published several discussion papers describing rules for the description of sequence changes that are currently widely accepted [Antonarakis et al., 1998;den Dunnen and Antonarakis, 2000].
Because of the importance of the issue and the overall consensus on the rules, Human Mutation is adopting an editorial policy that requests absolute compliance of these mutation nomenclature rules before manuscripts will be accepted and published.
A quick review across a range of journals that report sequence changes highlights the most offended rules (den Dunnen, in preparation). First, most papers fail to explicitly define which sequence file was used as a reference for numbering residues (nucleotides and amino acids). Consequently, a best guess is made, trying to