Skip to main content

Table 2 Study of model reproducibility: model learned from the T1D data set by switching training and test data subsets. Model learned from the T1D data set, with the same configuration as the model in Table 5 but switching training and test data subsets

From: A comparison of genomic profiles of complex diseases under different models

Chr Chr SNP Allele Allele Weight Genotypes Weight Genotypes
# Pos   1 2 rule 1 rule 1 rule 2 rule 2
6 32444658 SNP_A−1934589 A G 1.09 {0} 0.22 {1}
6 30135583 SNP_A−2111335 A G 0.74 {0}   
6 32444815 SNP_A−4303523 A G 0.62 {0, 1}   
6 30726039 SNP_A−2240847 A G 0.3 {0, 1, missing}   
19 39266932 SNP_A−4281637 A G 0.41 {0,1}   
6 31112694 SNP_A−4293786 C T 0.29 {0,1}   
6 31277959 SNP_A−1863445 A G 0.31 {0,1}   
6 31278044 SNP_A−1949560 A G 0.29 {0,1}   
1 113630788 SNP_A−2235405 A G 0.2 {0}   
  1. Weights and genotypes values are referred to as class 1, i.e. absence of disease