GenEx v.4.1.7 from MultiD

GenEx v4.1.7 from the Swedish company MultiD is a program used to analyze multivariate gene expression data. It has been primarily developed for analysis of real-time PCR data, but can be used for microarray data analysis as well. GenEx is the first comprehensive tool that was developed for multivariate analysis, such as clustering and principal component analysis of real-time PCR data. It has a strongly explorative character, utilizing visual images. GenEx has impressive user interface graphics as well as excellent plotting tools. The entire user interface is well-organized and relatively intuitive. The organization of sessions into projects, datasets and data files takes some getting used to, but provides the necessary adherence to regulatory requirements as far as the conservation of original data is concerned. There is a powerful preprocessing module with some dozen of data processing functions. Ct values can be normalized, translated into quantities, transformed by Logs or Ln and rescaled. Computations are performed as separate jobs on preprocessed data.

The GenEx software offers two separate approaches for analysis: Data can be classified as groups of genes or groups of samples. Genes that form a cluster have similar expression, while samples that are, for example, negative and positive for a disease should fall in different groups if proper expression markers are measured. To switch between classification of genes and classification of samples the data are transposed. The workflow of GenEx is reversible and several preprocessing and analytical steps can be performed interactively within one working session.

The computation apparatus of GenEx is robust and provides various multivariate approaches, whereas the most useful are probably the hierarchical clustering and principal component analysis. Both methods are based on a distance matrix of auto-scaled data with only minimal information loss but robust output. Both also provide excellent graphical images of gene associations or sample associations. GenEx contains also two, already well-established, algorithms for determination of the optimal normalization genes: The geNorm (developed at University of Gent) and the NormFinder (developed at Aarhus University Hospital). Separately testing different methods on real data showed good accord between them. All results also supported well-investigated biological phenomena.

Further advanced computation modules such as self-organizing maps, potential curves and 3-way analysis of the data are also available in the 4.1.7, providing state-of-the-art methods for advanced users.

The entire architecture of GenEx is based on individual dialogues for each task. This may sometimes create the impression of high complexity. A more homogeneous design of control elements would surely improve the user friendliness and reduce the need for introductory training. Some of the data pre-processing functions also leave great room for discussion and are based on assumptions. A good example of this is the transformation of Ct values to quantities utilizing the Ct of a single reaction; this is not a reliable method of quantification. However, it is to be mentioned, that since the analytical modules utilize robust algorithms, the pre-processing steps are less important in most cases.

To conclude: GenEx is a significant improvement to recent Excel-based tools in real-time PCR data exploration and analysis. So far, GenEx runs only under Windows®, however, it surely deserves to be available also for Mac users.