Segregation analysis of 159 soft tissue sarcoma kindreds: comparison of fixed and sequential sampling schemes

In this study we compared parameter estimates and model hypotheses in pedigree data collected by fixed sampling with estimates and hypotheses derived by sequential sampling. Employing a fixed sampling scheme, we previously analyzed data on relatives of 159 childhood sarcoma patients. We have now extracted from that data set individuals who would have been included in a sequentially sampled study. We applied segregation analysis to the truncated data, to determine the mode of inheritance and major locus parameter estimates. With data from both sampling schemes we made a family-by-family comparison to determine each family's contribution to a major gene model. The two sampling schemes yielded similar results: we detected segregation of a dominant major gene and obtained similar major locus parameter estimates. However, the sequential sampling scheme derived these conclusions from data on 982 relatives rather than the 2,451 ascertained in the fixed sampling scheme. The sequential sampling scheme failed to identify only one of the kindreds likely to be segregating the gene. For this data set, the sequential sampling scheme would have provided an efficient mechanism to discriminate genetic hypotheses and would have permitted focus of resources on the specific kindreds likely to segregate a major gene.