Modern biomedicine involves datasets with a large number of variables. For this reason, the Katiska software was developed to visualize complex patterns of interactions via a network approach. The software calculates the correlation matrix of a continuous dataset, and then uses the Himmeli tool to create pruned visualizations of the correlation structure.
The so-called Friedewald formula for LDL cholesterol has been at the heart of cardiovascular risk assessment since the 70’s, but recently the scientific community has focused on the finer structure of lipoprotein particles. Lipido is an epidemiologically validated software to extract more detailed information from the basic lipids (triglycerides, total and HDL cholesterol), without any new biochemical measurements.
The self-organizing map is a popular data mining tool, but so far it has not been fully applied outside engineering sciences. Melikerion is an open source package for SOM analyses in the Octave/Matlab programming environments, but you can also take advantage of the multi-dimensional visualization via the online service.
CraneFoot is an open-source pedigree drawing software that can process hundreds of families simultaneously and print a comprehensive report of the results. An online drawing service is provided for those that do not wish to use the software without a graphical user interface.
Drawing complex networks is no trivial matter especially if they are weighted, and for this reason an automated solution was created. Himmeli is built on a simulated annealing algorithm that spreads the nodes in two dimensions for best viewing on printed media.
Managing and preprocessing datasets is crucial for efficient and accurate statistical modelling. Most of the tasks are repetitive and generic in nature, and as such good targets for automation. The software tools in this site are designed primarily for the FinnDiane study group, but may be of use to other researchers as well.