Table 3: Regions with strongest imputation-inferred associations ($P<.0001$) using estimated allelic dosage under an adjusted additive genetic model.

Location SNP with best $P$-value within this region ${\text{Position}}^{*}$ Best $P$-${\text{value}}^{\S }$ # of top SNPs within this region Nearest $\text{gen}{\text{e}}^{*}$ 6p11-12 rs6909083 54290262 $1.80E-05$ 7 TINAG 6q22 rs17083119 121443809 $2.76E-05$ 4 C6orf170 1q23 rs1033465 171254353 $4.50E-05$ 2 TNFSF18 1p13 rs11583330 109925036 $5.35E-05$ 1 GNAI3 1q42 rs3014267 225619540 $6.58E-05$ 1 CDC42BPA 15q13 rs11635920 30999949 $7.18E-05$ 4 FMN1, GREM1 2q35-36 rs6726798 219009099 $8.66E-05$ 1 VIL1 10q21 rs11812882 59712631 $8.85E-05$ 1 ZCD1 2q35 rs1106412 219023301 $8.91E-05$ 1 USP37 3p24 rs11927173 23200198 $9.39E-05$ 2 UBE2E2 8q22 rs3098241 104494480 $9.72E-05$ 2 SLC25A32

${}^{*}$Affymetrix NetAffx annotation build 25, NCBI genome build 36.1 ${}^{\S }P$-value of logistic regression under additive genetic model, adjusted for age, gender, diabetes duration, and HbA1C level.