Introduction

Reverse-phase protein arrays (RPPA) provide an effective means for the high-throughput profiling of protein expression in large sets of samples. Thus, this technology is typically applied in conjunction with complex factorial study designs, where differential protein expression and modification shall be monitored over time under different experimental conditions, such as the exposure to diverse compounds administered at different dose levels. While a wide variety of bioinformatics tools is available for the most common microarray platform types (e.g., gene-centered arrays, SNP arrays, exon arrays, tiling arrays, etc.), the current repertoire of specialized analysis and visualization tools accounting for the specific characteristics of RPPA datasets is scarce. In order to fill this gap, we developed the tool RPPApipe, which offers various data processing functions, which have been specifically designed for the analysis of RPPA data. The tool runs in any current browser and is available from our Bioinformatics Toolbox.

After uploading in a tabular format which is specially geared to RPPA data, our web-based analysis pipeline provides diverse methods for data transformation and missing value imputation as well as for protein annotation and ID conversion. Along with the expression data, the user can also upload design information. Currently, we support three types of experimental designs: 'Paired sample groups' (e.g., treatments and corresponding controls or mutant vs. wild-type), 'Multiple sample groups' (e.g., different tissues or cell types) and 'Replicated time-series data' (e.g., repeated dose studies). The third type corresponds to a study design, where time course measurements shall be compared and multiple replicates are available for each time point. Depending on your experimental design, appropriate methods for statistical analysis are proposed. Differential protein expression can be visualized in volcano plots and venn diagrams. One of the unique features offered by RPPApipe is the visualization of differential protein modification, i.e., differential expression between the unmodified and the modified form of a certain protein, by means of specially adapted volcano plots. In addition, the alteration of canonical signaling pathways can be inspected by generating purpose-built pathway profile diagrams. Another unique feature offered by RPPApipe is the simplification of RPPA datasets by representating the complex time-resolved expression profiles observed for a certain protein by abstracted, discrete regulation states (see documentation). The user can select if abstracted regulation states or fold-changes shall be used for the clustering of the data which can be visualized as a specially adjusted heatmap with dendrograms. All plots can be downloaded as publication quality vector graphics in PDF format. For convenience, the user can use predefined workflows or create custom-built ones for RPPA data processing. Finally, the results from the statistical analysis can be exported to our pathway analysis tool InCroMAP which offers specific functions for visual inspection of the data in the context of relevant signaling pathways.

In the following you can find a collection of plots which were generated using the RPPApipe online tool. You can click on a preview picture to enlarge it.