DepthTools: an R package for a robust analysis of gene expression data

Abstract

Background: The use of DNA microarrays and oligonucleotide chips of high density in modern biomedical research provides complex, high dimensional data which have been proven to convey crucial information about gene expression levels and to play an important role in disease diagnosis. Therefore, there is a need for developing new, robust statistical techniques to analyze these data.

Results: depthTools is an R package for a robust statistical analysis of gene expression data, based on an efficient implementation of a feasible notion of depth, the Modified Band Depth. This software includes several visualization and inference tools successfully applied to high dimensional gene expression data. A user-friendly interface is also provided via an R-commander plugin.

Conclusion: We illustrate the utility of the depthTools package, that could be used, for instance, to achieve a better understanding of genome-level variation between tumors and to facilitate the development of personalized treatments.

Figure 1

MBD plots. (a) Representation in parallel coordinates of 25 normal prostate samples, with the deepest one depicted in red. (b) MBD-based bands, for different proportions of central points (grayscale regions), corresponding to 25 normal prostate samples with respect to 25 cancer prostate samples (blue lines). The normal sample which lies the deepest in the reference collection of tumor ones is drawn in red.

Figure 2

Trimmed mean plots. (a) Representation in parallel coordinates of the 0.25-trimmed mean of the normal prostate samples, in black. The trimmed out 25% most external points are depicted in gray; the remaining samples, used to compute the trimmed mean, are drawn in blue. (b) Trimmed means for different proportions of trimmed out points, corresponding to the normal (blue-gray) and cancer (red-gray) samples, for a subset of genes.

Figure 3

Plot of the 25% most central normal prostate samples. Representation in parallel coordinates of the normal samples in the prostate dataset with the 7 most central ones (25%) represented with solid lines and coloured from center outwards with a red-yellow palette; the remaining samples are shown as gray, dotted lines.

Figure 4

Scale curves of normal and tumor prostate samples. Scale curves for the normal and tumor samples included in the prostate data. The tumor samples (dashed line) have in general a larger dispersion than the normal ones (solid line). The red dot represents the spread of the 25% most central normal samples.

Figure 5

Depth tools menu. All the functions included in the depthTools package are available through the menu inserted in the R-commander bar.

Figure 6

MBD windows. (a) Main window for the MBD computation in the R-commander. (b) Graphical window for adjusting the appearance of the MBD plot. (c) Output window for selecting which computations are stored as R objects.