Exploring linkage of chromosome 18 markers and bipolar disease.

The linkage reports of bipolar disease and chromosome 18 markers are controversial. We used the GAW10 data sets to further explore several observations: 1) a possible parent-of-origin effect with only 'paternal' pedigrees showing linkage; 2) the preponderance of women affected with bipolar disease, and 3) the possible existence of phenocopies in the bipolar data sets. We performed linkage analysis allowing for independent male/female recombination fractions. Our hypothesis was that if there is linkage only in 'paternal' pedigrees, then the lod score would maximize at low male and high female recombination fractions. We did not find such an effect in the combined data set. There was no consistent effect on the difference of male and female recombination fractions, suggesting that an effect is not detectable in this data set with this method. In addition, there is interesting evidence for a recessively inherited, highly penetrant gene in a subset of families. Allowing for higher penetrances for bipolar disease in women than in men had no effect on the lod scores. There was also not much difference in the lod scores calculated under the assumption of a phenocopy rate versus no phenocopies. From simulation studies, we would have expected some effect if there were linkage and phenocopies were present.