No association survived correction for multiple testing in this study, though three variants in GCH1 did show significant association before correction and would merit follow-up in a larger case-control cohort.
Thus, even though multiple testing of array data indicated lack of statistical significance for the altered probes, our single gene validation experiments suggest that HIPPI indeed interferes with these processes in cell.
This threshold is arbitrary and lacks multiple test correction, though our goal is to produce a list of potential candidates that is inclusive, though it may include some false positives.
We also found significant classification in PMv, SMA, and pre-SMA, each of which was highly significant after correcting for multiple tests, even though the classification accuracies were somewhat lower in these regions.
P-values smaller than 0.05 were considered to indicate statistical significance though no correction for multiple testing was applied.
Moreover, tests for multiplicative allelic trend (genotypes coded 0, 1, 2) also showed that none was significant at the 5% level with, even though no correction for multiple testing was applied.
Second, even though we corrected for multiple testing, our results could represent false-positive findings and should be considered hypothesis generating.
GWA studies use a conservative threshold for significance that corrects for genome-wide multiple testing (usually p < 5 × 10−8), even though only relatively large effect sizes can clear such a high hurdle.
Even though we had corrected for multiple testing, this approach may have been ineffective to correct for false positives.
Some might be false positives, even though P-values were corrected for multiple testing.
