Statistical assessment for PLS is done using permutation tests for the LVs and bootstrap estimation of standard errors for the voxel saliences.
After normalization, only those genes with a detection P-value <0.01 in all samples (Illumina Inc., 2009) or considered present in all samples according to the Affymetrix MAS5 algorithm (Affymetrix Inc., 2009) were included in further analyses (i.e. bootstrap estimation of p A by the procedure described in Section 3.4).
Consistent with present results, they found that both estimators were equally efficient (based on their results from bootstrap estimation of standard errors) and both were appreciably more efficient than results obtained using simple random sampling by at least 24%, rising to as much as 61% as sample size increased.
Inter-rater reliability was estimated with the weighted and unweighted Kappa statistic of nominal categorical and ordered categorical variables respectively, and 95% confidence intervals were calculated using bootstrap estimation of standard error for Kappa.
ML tree searches and bootstrap estimation of clade support were conducted with RAxML [ 53].
Bootstrap estimations of 95% confidence intervals for the averages and medians are also shown.
Moreover, such percentiles have low variance, low dependence on sample size, and are amenable to bootstrap estimations of confidence intervals, when compared to the largest individual per sample [ 108].
To reduce this, the use of unitary weighting and bootstrap estimations of confidence intervals are presented.
Confidence intervals in the data were not available, so an optimistic 3% relative and 0.5% absolute error was assumed for each data point (σ ij = 0. 03 x ^ i (t j ) + 0. 005 max (x ^ i ) ). Figure 5A shows the resulting time-series profiles for bootstrap estimations of each model.
High-content image-profiling, including the PCA, Jonckheere-Terpstra test, Pearson product-moment correlation analysis, and bootstrap-based estimation of the FDR [17], was performed using a newly built Java program.
