Figure 1.
Description of the stratification protocol used in this study.
The 4 original sub-phenotypes (denoted by four different colors) were further stratified by removal of families containing affected siblings of another sub-phenotype to yield the Gs level subgroups. These subgroups were further divided according to male only or female-containing pedigrees. Due to intra-family heterogeneity in multiplex cases, some families were included in more than one stratified group. Therefore, the sum of individual and family numbers in subgroups exceeds the numbers listed for the original combined cohort (ALL).
Table 1.
Improvement of maximum LOD scores in subgroups with addition of new families.
Table 2.
Loci with highest LOD scores for a given subtype.
Table 3.
Linkage data obtained for four overlapping regions, between two different subgroups.
Figure 2.
A graphical representation (heat map) of the LOD score data (cut-off ≥2.0) was generated to visually demonstrate the computed linkage scores for each subgroup in a hierarchy. The heat map compares LOD score patterns for the 16 subgroups. As expected, there were more similarities within each ADI-R group (e.g., G1, G1s, G1M, and G1Fc) than between two different ADI-R groups. Each horizontal band represents a SNP while each column represents a stratified subgroup, with the exception of the first column which represents the combined (ALL) cohort. Table S4 in File S2 lists the SNPs and LOD scores contributing to the identified segregation patterns by subgroups (i.e., hot spots). The corresponding genomic positions of the SNPs contributed to the heat map (Y-axis) are listed in Table S4 in File S2. The heat map was generated using MeV software [41].