R environment

All statistical analyses were done in R environment (www.r-project.org). Trehalose and glycogen levels, as well as OCRs were compared with analysis of variance (ANOVA) followed by Tukey’s honestly significant differences (HSD) post-hoc test. Trehalose/glycogen amounts were log-transformed prior to model fit, normality was confirmed with QQ-plots and Shapiro-Wilk test, homoscedasticity with Levene’s test. Survival rates after desiccation and rehydration were compared with beta regression as described before (Erkut et al., 2013 (link)), followed by Type II analysis of deviance for generalized linear models. Prior to beta regression, fit to beta distribution was confirmed with QQ-plots. Statistical power was calculated via power analysis when possible. The maximum Type I error rate was set as α = 0.05 for all tests. Data are presented as mean ± standard error for C. elegans trehalose levels and survival rates, and mean ± 95% confidence limit for other measurements unless stated otherwise.

Statistical analysis and PCA was done using the R environment (https://www.r-project.org/). Differences between two groups were analyzed using an unpaired, two-tailed Student’s T-test. In parallel the samples were tested for significant variation of variance, and if necessary a Welch correction was included in the statistical analysis. All statistical tests were performed between sets of individual biological replicates.

We used the R environment (http://www.r-project.org/) for statistical analysis. For univariate analysis, we chose to apply a continuous regression analysis (Spearman correlation) for each feature relative to the GA at time of sampling. Shaded areas on linear regression graphs represent 95% confidence intervals. Multiple comparison corrections were performed using the Benjamini-Hochberg procedure and reported as false discovery rates (FDR). For the multivariate analysis, an Elastic Net (EN) regularized regression model was trained on the pre-processed dataset using the R package glmnet (version 4.0-2). In the EN model, the immune features derived from mass cytometric analysis were used to predict GA, which was treated as a continuous variable. An underlying assumption of the EN algorithm is statistical independence between all observations. In this analysis, we assume independence as the samples come from different subjects.

All statistical analyses were done in the R environment (R Foundation for Statistical Computing). A 2-tailed Student t test was applied for each feature to determine significant differences of enrichment between the two populations, and the resulting P values were adjusted with Bonferroni multiple testing correction. Adjusted P values below .01 were considered significantly different.

All analyses were performed in the R environment (version 3.4.0, R Foundation for Statistical Computing, Vienna, Austria). The volumetric analysis was conducted firstly on the four prior ROIs (amygdala, hippocampus, ACC and PFC) and then on the brain regions in Neuromorphometrics parcellation for exploratory purposes. The distribution of the volumes was assessed for normality using the Shapiro–Wilk test. Multivariate analyses of variance (MANOVA) were conducted on volumetric measures of brain regions using a two-factor design, where group was the between-participant factor, hemisphere was the within-participant factor, and age was the covariate. To further examine the relationship between brain structures and PTSD symptom severity, partial correlations controlling for age were calculated between the analysed measures and CAPS-5 severity scores in the group of participants diagnosed with PTSD. A significance level of P < 0.05 was used for all analyses. No corrections were applied for unequal group sizes.

The log 2 ratios of the samples were compared using permutation t-tests at simulation numbers of 100,000, adjusted for age and sex. An adjusted p value with false discovery rate correction less than 0.05 was considered significant. To find the DEGs with a linear association with increasing T, N stage and perineural invasion, DEGs that had positive coefficient were selected for further analysis. To evaluate the clinical effect of ceruloplasmin expression, we compared clinicopathological characteristics with the χ²-test. A p value of less than 0.05 was considered significant. The Kaplan-Meier method was used to calculate survival rates for candidate genes, compared using the log-rank test. All statistical analyses were performed in the R environment (The R Foundation for Statistical Computing, Vienna, Austria) and IBM SPSS Statistics version 21.0 (IBM Corp., Somers, NY, USA), with a p value less than 0.05 considered significant.

The cumulative incidence of pneumonia was estimated using the Kaplan–Meier method and log-rank test. A p-value of <0.05 was considered statistically significant. Recursive partitioning was performed by the CART algorithm in R [7 ,8 ]. Statistical analyses were performed using GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA), JMP^® 13 (SAS Institute, Cary, NC, USA) and the R environment (version 4.1.1, R Foundation, Vienna, Austria) [9 ].