Fig 1.
Study design (a) and experimental procedure (b).
Table 1.
Sample characteristics at baseline.
Fig 2.
Placebo effects on nausea ratings, motion sickness, and normo-to-tachy ratio in the electrogastrogram.
Boxplots depict day-adjusted scores (DAS) for the three nausea measures: (a) Nausea (DAS-Nausea), (b) motion sickness (DAS-MS), and (c) normo-to-tachy ratio in the electrogastrogram (DAS-NTT). DAS-Nausea, Fgroup(1,86) = 44.83, p < 0.001; DAS-MS, Fgroup(1,84) = 14.93, p < 0.001; DAS-NTT, Fint(1,83) = 4.16, p = 0.044 (female, Fgroup(1,40) = 4.10, p = 0.049; male, Fgroup(1,43) = 0.83, p = 0.366); See S1 Table for further details. On each box, the central dot indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points not considered outliers, and the outliers are plotted as black dots.
Fig 3.
Identified plasma proteins and those significantly affected by the placebo treatment.
(a) The heatmap depicts the sample (column) vs protein (row) matrix with hierarchical clustering dendrograms. Abundances were row-wise z-score transformed. The sample features group, experiment day, time point of measurement, sex and age are color-/bar-coded at the bottom of the heatmap. (b) Heatmap of the 74 proteins identified as significantly affected (p < 0.05) by placebo treatment using ANCOVA, with fold changes on Day 1 included as covariates. Numbers on the left indicate numbers of peptides associated with the protein. Color refers to average log-ratio (measurement 2 vs measurement 1) of all protein-associated peptides. Columns are labeled with the same color-code as in A. Multiple protein labels arise from non-uniquely mapped peptides.
Fig 4.
Variance composition of protein changes on Day 2.
(a) The amount of variance explained was estimated using linear multiple regression models to predict protein fold change variance by different experimental factors (DAS, sex, group) and their interactions (predictor variables). Models were generated independently for each protein and type of nausea measure: Nausea, NTT and MS. Barplot histograms depict variance composition for all proteins significantly affected (p < 0.05) by at least one predictor variable. Multiple protein labels arise from non-uniquely mapped peptides. (b) GO enrichment for each group of significantly regulated proteins (p < 0.05). Dot color and size refer to FDR-corrected enrichment -log10 (p-value).
Fig 5.
StringDB network and Gene Ontology (GO) enrichment analysis of placebo-related proteins.
(a) StringDB network of placebo-related proteins. Nodes are colored according to Gene Ontology (GO) term group association. Numbers refer to functional clusters. GO functional clusters were created based on shared members (proteins). Edges refer to StringDB interactions (see methods). (b) Expression patterns for selected GO term proteins. Barplots depict average log expression (measurement 2 versus measurement 1) of all protein-associated peptides. Barplot colors refer to day and group. Error bars denote standard error of the mean. Boxes around barplots indicate GO term membership, and the color of the boxes refer to GO functional cluster associated in panel A. (c) The fitted GO term-based linear multiple regression models to predict DAS. Acute-phase response proteins predict DAS-Nausea in control (top) and placebo group (bottom); complement activation proteins predict DAS-NTT in control and placebo groups; hydrogen peroxide catabolic process proteins predict DAS-MS in control and placebo groups; Model p-values (FDR-corrected) and model R-squared are specified in the plots. Blue dots are model outputs for each data point. The linear fit and 95% confident bands are denoted by solid and dashed red lines.
Fig 6.
Prediction of placebo responders by Day 2 baseline proteins.
(a) Adapted confusion matrices for placebo responders vs. non-responders in the placebo group for nausea and motion sickness. Placebo responders were defined as participants showing a reduction of at least 50% in day-adjusted scores for nausea (DAS-Nausea) and motion sickness (DAS-MS), respectively. (b) Receiver operating characteristic (ROC) curves for the support vector machine (SVM) models. The blue line refers to the ‘ANOVA plus model’. Yellow area refers to the range of all 10 ‘RANDOM model’ permutations, area under the curve (AUC) values in mean and standard deviation given.