normalized using BeadStudio software (Illumina, San Diego, CA, USA). All subjects had FKBP5 expression values above a threshold detection level of p<0.00001. RNA expression was analyzed using mixed models as described above for genotype. To validate the array-based FKBP5 expression values we used RT-PCR to confirm the association with cortisol levels. RNA from 24 subjects (from 4 different arrays) was reverse transcribed using the ABI High Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's specifications. One-fortieth of each cDNA reaction was amplified in triplicate in a 10 μl PCR reaction containing 500 nM forward and 500 nM reverse primer, 0.1 × Sybr Green I (Molecular Probes, Eugene, OR, USA), and 1x TaqMan Genotyping Master Mix (Applied Biosystems, Foster City, CA, USA) using an ABI Prism 7900HT Sequence Detection System. The expression of a housekeeping gene, HPRT, was used to normalize expression values between subjects. Primer sequences were, HPRT Fwd TGAGGATTTGGAAAGGGTGT, HPRT Rev CCTCCCATCTCCTTCATCAC, FKBP5 Fwd AAGTTTGCAGAGCAGGATGC, FKBP5 Rev GGCCCTCAGGTTTCTCTTCT. The relative level of gene expression was determined with the ΔΔCt method using the mean value for two experiments. Consistent with array-based results, the mean cycle threshold was associated with cortisol levels at time point
