Open Mx is a statistical modeling program which is applicable in levels of scientific scope from the genomic to individual behavior and social interactions, all the way up to the national and state epidemiological data. Nested statistical models are essential to disentangle the effects of one level of scope from the next. In order to prepare Open Mx for the statistical challenges of the coming decades, the software will take advantage of parallel computing resources so that computationally intensive statistical problems can be executed significantly faster on large scale computing grids. The primary aim of the Open Mx project is to build a statistical software tool that enables and accelerates the pace of funded research in the social, behavioral and medical sciences.

Structural Equation Modeling (SEM) is used by diverse set of health-relevant disciplines including genetic and non-genetic studies of addictive behavior, psychopathology, cardiovascular disease and cancer research. Frequently, studies are faced with very large datasets; this is the case for neuroimaging, genome-wide association, and electrophysiology or other time-varying aspects of human individual differences. At the same time, the measurement of complex traits is often difficult, which creates a further challenge to their statistical analysis. The challenges of large data sets and complex traits are shared by projects at all levels of scientific scope. The Open Mx software will address many of these data analytic needs in a free, open source and extensible software package that will run on operating systems including Linux, Apple OS X, and Windows.

OpenMx is expected to enter open beta in the fall of 2009.

More information can be found at the OpenMx SEM user’s web site.