Life Science students (and Professors) need to perform multiple simple regression analytics comparing a large number of dependent variables to a single independent variable (a "control" vector). Yes, this is simplistic in the true sense, but it is a late step in which a large number of samples are quantified, merged, normalized, and finally compared for "statistically significant differences". These are not binary comparisons but usually detect differences based on linear regression slope and correlation comparisons. A p-value is required for publications. The best example of such a software I can locate is the "Daniels XL-Toolbox" plugin (http://xltoolbox.sourceforge.net/) by Daniel Kraus at Harvard University. 

This has been a long-standing deficiency with Excel, and if Gnumeric could supply this capability, it would improve the lives of thousands of people in higher education, that lack moderate statistical backgrounds. This is particularly true in Biology, where a deluge of many petabytes of data are expected per year in DNA sequencing technologies. This sequencing technology will extend to environmental diversity, Phylogenetics, agricultural breeding, and clinical personalized medicine.

Thanks, 

Peter