Professor Crow takes issue with our view that DISC1 is important to schizophrenia in general and is not restricted to the initial family in which disruption of this gene was reported. His argument is based on a selected set of sibpair studies whose results do not support linkage anywhere on chromosome 1. This finding was unsurprising in view of the lack of power of such studies in the presence of genetic heterogeneity in schizophrenia susceptibility, which was not mentioned by Professor Crow. We and a large number of other workers in the field consider that such locus heterogeneity is highly likely and have shown that the sib-pair strategy has limited power to detect a locus that contributes less than 20% of the variance (Reference Macgregor, Visscher and KnottMacgregor et al, 2002). Where heterogeneity is expected then linkage analysis, especially of extended multiplex pedigrees, and gene candidacy identified though the investigation of psychosis-associated karyotype anomalies are appropriate research strategies. Where there is a priori evidence from cytogenetic and linkage studies (such as the Finnish studies mentioned in the editorial) then the case–control association approach provides a useful resource to delineate potential population haplotype distortions that may indicate underlying functional mutations.

We would therefore disagree strongly with Crow in his statement that we have ‘overinterpreted’ the importance of DISC1 and commend an excellent review of schizophrenia neurobiology which emphasises heterogeneity (Reference Ross, Margolis and ReadingRoss et al, 2006). Although our theoretical framework differs from that of Bleuler (Reference Bleuler and Zinkin1950), we feel that the recent genetics and neurological discoveries are in agreement with his position that there is indeed a ‘group of schizophrenias’.