Skip to main content
. Author manuscript; available in PMC: 2012 Sep 28.
Published in final edited form as: Biometrics. 2011 Dec 7;68(3):766–773. doi: 10.1111/j.1541-0420.2011.01713.x

Table 4.

Average computing times (in seconds) over 10 (in the case of gene–gene interactions, 5) applications of the gTDT based on both analytic (or numerical) parameter estimates using log-likelihoods of the form (2) and the conventional iterative fitting procedure as well as score tests based on (2) to the 569,187 autosomal SNPs from the case–parent trio study of the International Cleft Consortium (Beaty et al., 2010), interactions of each of these SNPs with gender, and all 499,500 interactions between two of 1000 randomly selected SNPs (considering a model consisting of one parameter for the interaction term)

Individual SNPs
Additive Dominant Recessive MAX
Analytic 525 474 472 690
Score 525 437 435 698
Conventional 43,318 42,740 42,726 128,784

SNP–Gender Interaction
Additive Dominant Recessive

Analytic 545 455 454
Score 556 465 462
Conventional 51,600 51,834 53,237

SNP–SNP Interaction
Additive Dominant Recessive

Analytic 335
Numerical 1,182 497 469
Score 785 324 283
Conventional 126,748 124,753 124,425