Improved evidence for linkage on 6p and 5p with retrospective pooling of data from three asthma genome screens

To improve our ability to identify genes for complex diseases, evaluation of new methods that retrospectively pool genotype and phenotype data collected by multiple centers is important. Availability of three whole-genome screens enabled us to compare two methods, pooling raw data and meta-analysis. Multipoint linkage analyses were performed on two outcomes, total serum IgE levels and asthma affection status, using an improved Haseman-Elston algorithm. Two regions showed stronger evidence for linkage using covariate-adjusted pooled data, compared with any individual sample. Both methods for pooling data identified strong linkage to Z-transformed logeIgE levels at a location between D6S1019 and D6S426, and to the asthma trait at D5S268. In conclusion, retrospective analysis of pooled genome scan data is a potentially powerful and useful method to examine both positive and negative evidence for linkage of quantitative and categorical phenotypes across populations.