Although we have sequenced most of the protein-coding exons of the X chromosome, the abnormal genes responsible for disease in most of the XLMR-affected families examined have not been identified. Of the subjects with XLMR screened, the likely genetic basis of 53 (25%) has now been established by detection of truncating, missense, in-frame or copy number variants found in this screen or by others analyzing the same cases in parallel (Supplementary Table 7 online). There are several plausible reasons for this. A proportion of X-chromosome genes were intractable to PCR primer design and, even when screened, coverage may not be complete. Moreover, a subset of disease-causing variants usually are in promoter regions, introns and other noncoding sequences, which we have not examined. Other classes of abnormality, for example copy number changes, inversions and other chromosomal rearrangements, are not detectable by a PCR-based resequencing approach. Indeed, some protein-coding genes may not yet have been annotated and it is conceivable that some XLMR-associated genes are not protein coding. It is also possible that mental retardation in a proportion of the families in
