Table 3. Nested ordered-subset analysis in the subgroup with early-onset schizophrenia by CPT or WCST.
Change in LOD | Maximum subset-based | Mean family covariate value (SE) | ||||||
Covariates | Cytogenetic location | Markers | (permutation p value)a | Initial LOD | LOD (genome-wide empirical p value)b | No. Families used/total (%) | Within nested subset | Remaining families |
CPT-‘1-9’ indices | ||||||||
Undegraded hit rate | 2q22.1 | D2S442 | 0.23 (0.54) | 4.23 | 4.46 | 275/282 (97.5) | --- | --- |
Undegraded false alarm rate | 2q22.1 | D2S442 | 3.13 (0.004) | 4.23 | 7.36 (<0.001) | 228/282 (80.9) | 1.84 (0.09)* | -0.04 (0.03) |
Undegraded d′ | 2q22.1 | D2S442 | 0.85 (0.08) | 4.23 | 5.08 | 206/282 (73.0) | --- | --- |
Undegraded reaction time | 2q22.1 | D2S442 | 0.48 (0.38) | 4.23 | 4.71 | 268/282 (95.0) | --- | --- |
Degraded hit rate | 2q22.1 | D2S442 | 0.70 (0.16) | 4.23 | 4.93 | 242/281 (86.1) | --- | --- |
Degraded false alarm rate | 2q22.1 | D2S442 | 3.48 (0.001) | 4.23 | 7.71 (<0.001) | 243/281 (86.5) | 1.49 (0.09)* | -0.26 (0.04) |
Degraded d′ | 2q22.1 | D2S442 | 1.06 (0.09) | 4.23 | 5.29 | 251/281 (89.3) | --- | --- |
Degraded reaction time | 2q22.1 | D2S442 | 1.13 (0.07) | 4.23 | 5.36 | 262/281 (93.2) | --- | --- |
WCST indices | ||||||||
Total Errors | 2q14.1 | D2S410 | 0.53 (0.08) | 3.84 | 4.37 | 67/263 (25.5) | --- | --- |
Non-perseverative Errors | 2q22.1 | D2S442 | 0.68 (0.16) | 3.64 | 4.32 | 247/263 (93.9) | --- | --- |
Perseverative Errors | 2q22.1 | D2S442 | 0.00 (1.00) | 3.64 | 4.17 | 263/263 (100.0) | --- | --- |
Perseverative Response | 2q22.1 | D2S442 | 0.00 (1.00) | 3.84 | 3.84 | 263/263 (100.0) | --- | --- |
Categories Achieved | 2q14.1 | D2S410 | 0.56 (0.34) | 3.18 | 3.74 | 55/263 (20.9) | --- | --- |
Conceptual Level Response | 2q14.1 | D2S410 | 0.41 (0.40) | 3.18 | 3.59 | 72/263 (27.4) | --- | --- |
Failure to Maintain Set | 2q22.1 | D2S442 | 0.69 (0.47) | 3.64 | 3.87 | 255/263 (97.0) | --- | --- |
Trials to Complete First Category | 2q22.1 | D2S442 | 0.76 (0.15) | 3.64 | 4.40 | 53/263 (20.2) | --- | --- |
Families were randomly permuted for 1000 times with respect to the covariate ranking and a chromosome-wide p value for each chromosome was yielded.
Significance level derived from simulations; a genome-wide empirical p-value <0.0015 [i.e., 0.05/33 covariates)] is denoted in boldface as reaching genome-wide significance.
*P<0.0001 for the mixed effect model comparing the families covariate values between the nested subset of families and the remaining ones.