Simca 16

Principal component analysis (PCA) and partial least squares discriminant analysis (OPLS-DA) was performed in SIMCA 16 (Sartorius Stedim Data Analytics AB, Umeå, Sweden) using the normalized, filtered data. Unit Variance (UV) scaling was used for all multivariate plots. PCA plots were used to assess data quality by verifying the clustering and centering of QCSP samples by PCA, and OPLS-DA plots were used to assess the separation of metabolomes between vehicle and treated cells, as well as to calculate variable importance to projection (VIP) scores for each peak. Heatmaps were generated using MetaboAnalyst 5.0 [39 (link)]. Fold changes and p-values were calculated for each peak for each treatment as compared to the vehicle control. p-values were calculated using Student’s t-test. Correlation analyses were performed by using the Statistical Analysis (metadata table) module in MetaboAnalyst 5.0 using Pearson r as the correlation measure. p-values were not adjusted for multiple testing due to the small sample size of this study and the exploratory, rather than confirmatory, nature of this study [40 (link)]. Pathway analyses were conducted using the “Functional Analysis” module of MetaboAnalyst 5.0 using all peaks in the metabolomics dataset. Metabolites were mapped on the KEGG metabolite set library using a p-value cutoff of 0.05.

Mean expression intensities of markers for all subpopulations on all mice of the study were pooled in one single spreadsheet per R-phenograph cluster per organ, transformed in asinh and centered to the mean. The resulting list was run in the SIMCA16 multivariate analysis software (Sartorius). Orthogonal partial least square-discriminant analysis (OPLS-DA) method was applied on the dataset to identify groups of samples presenting similar types of variations. Variable importance in projection (VIP) was selected with a VIP value over 1. A list of parameters with VIP >1 was subjected to reduction. Predictive variables with VIP >1 were uploaded into TMeV, a microarray software suite, in order to build a heatmap after median centering and SD reduction. Dataset was clustered using Euclidian distance on samples and variables by hierarchical clustering. All independent values were visualized with a color code spanning values of −1 to +1.

Unless noted otherwise, one-way ANOVA or two-way ANOVA was used with post-hoc comparisons in Prism9. Principal component analysis was performed using SIMCA 16 (Sartorius, Göttingen, Germany). Pearson correlation coefficients were calculated for each pairwise combination of samples using the Pearson function from the scipy.stats module in Python. These results were visualized with Matplotlib’s matshow function. For far-close tumor–macrophage distance comparison, one-tailed t-test was performed to compare the average speed of close and far macrophages.

Baseline and phase corrections were conducted for raw spectra of intestinal contents and feces in Topspin 3.6.2 before processing in Matlab 2018b. Spectral processing followed our previous method.^[22] Multivariate data analysis (MVDA) was performed on the processed NMR spectra using SIMCA 16 (Sartorius, Umeå, Sweden). MVDA included unsupervised principal component analysis (PCA) and supervised orthogonal projections to latent structures discriminant analysis (OPLS‐DA). For OPLS‐DA models, a cross‐validation procedure using venetian blinds with seven segments was conducted. The S‐line plot of the OPLS‐DA models was used to visualize the differences in spectral signal intensity between two groups. The S‐line plot also reveals the correlations between absolute values of variables and predictive scores (p(corr)) by the color. p(corr)>0.6 indicates that a variable is important to the group discrimination.^[43] For the quantification of metabolites, Chenomx (Version 8.6, Chenomx Inc., Edmonton, Alberta, Canada) was used.

Results of impedometric measurements were statistically analyzed with a two-way ANOVA model performed using PRC GLM of SAS (SAS Inst. Inc., Cary, NC, USA), whereas SIMCA 16 (Sartorius Stedim Data Analytics, Gottinga, Germany) software was used to create a principal component analysis (PCA) biplot to get visual interpretation of the data analyzed.

The normalized, filtered data were imported into SIMCA 16 (Sartorius Stedim Data Analytics AB, Umeå, Sweden), scaled using Unit Variance (UV) scaling, and then used to generate principal component analysis (PCA) and orthogonal partial least squares-discriminant analysis (OPLS-DA). PCA plots were used to assess data quality and clustering of QCSP samples, and OPLS-DA plots were used to assess the separation of metabolomes between vehicle and treated cells as well as to calculate variable importance to projection (VIP) scores for each peak. Heatmaps were generated using MetaboAnalyst 5.0. Fold changes and p-values were calculated for each peak for each treatment as compared to the vehicle control. p-values were calculated using Student’s t-test. p-values were not adjusted for multiple testing due to the small sample size of this study and the exploratory, rather than confirmatory, nature of this study [100 (link)].

Analysis of each experiment was carried out in triplicate and all data were represented as the mean SD. The principal component analysis (PCA) was performed using the mean data with SIMCA 16.0.2 statistical software (Umetrics Inc., San Jose, CA, USA). All the data were normalized using log transformation to have a normal distribution.