Bipolar disorder: The correlation of 3 analytic results for a genome-wide parent of origin effect

We have studied 71 multiplex bipolar pedigrees ascertained through a treated BPI proband and 2 affected first degree relatives. Six families did not meet strict ascertainment criteria. In the 65 remaining pedigrees, there were 583 interviewed and genotyped subjects; 129 bipolar I (BPI), 99 bipolar II (BPII), 71 recurrent unipolar depression (RUP), and 7 schizoaffective-manic (SA-M). We analyzed the total sample and then partitioned accordingto the sex of the affected parent. There were 23 "paternal" and 34 "maternal" pedigrees; 8 were unclassified. Analytic methods included multipoint parametric and nonparametric analyses (GENEHUNTER) and IBD analyses (single point) of affected sib pair (SIBPAL) and relative pair (simIBD). Our results suggested strongest evidence for susceptibility loci in 4 regions: at 2p13, 8q24, 13q23 and 18q21-23. While the increased allele sharing on 2p and 18q were primarily in the paternal pedigrees, the findings on 8q and 13q were primarily in the maternal pedigrees. The results of the three genetic analytic programs were correlated with each other asing the inverse of the natural log of the p value. When comparing the results of the different analyses in the total sample of 65 pedigrees, the Pearson correlation coefficient ranged from 0.32 to 0.66. The maternal and paternal pedigrees were tested to determine if there was correlation between these two groups. Increased IBD sharing in maternal pedigrees did not predict increased sharing in paternal pedigrees. In fact, there was a slightly negative (but significant) correlation (r = -0.11) in the mean sharing comparison of the two groups, suggesting that increased sharing in paternal pedigrees predicted decreased sharing in maternal pedigrees. These analyses provide further evidence for a genome-wide parent of origin effect in bipolar disorder.