Rater reliability of fragile X mutation size estimates: a multilaboratory analysis

Notwithstanding the use of comparable molecular protocols, description and measurement of the fra(X) (fragile X) mutation may vary according to its appearance as a discrete band, smear, multiple bands, or mosaic. Estimation of mutation size may also differ from one laboratory to another. We report on the description of an mutation size estimate for a large sample of individuals tested for the fra(X) pre- or full mutation. Of 63 DNA samples evaluated, 45 were identified previously as fra(X) pre- or full mutations. DNA from 18 unaffected individuals was used as control. Genomic DNA was extracted from peripheral blood, and DNA fragments from each of four laboratories were sent to a single center where Southern blots were prepared and hybridized with the pE5.1 probe. Photographs from autoradiographs were returned to each site, and raters blind to the identity of the specimens were asked to evaluate them. Raters' estimates of mutation size compared favorably with a reference test. Intrarater reliability was good to excellent. Variability in mutation size estimates was comparable across band types. Variability in estimates was moderate, and was significantly correlated with absolute mutation size and band type.