Optimization of Computations for Structural Equation Modeling with Applications in Bionformatics

Structural equation modeling (SEM) is a technique for analysis of linear relations represented as the causal and correlational relationships between observed and latent variables. SEM is a popular tool in a wide range of fields, from the humanities to the natural sciences. Over the past decade, this method has become especially interesting in areas that are at the interface with biology. However, the common assumption that observations are independent is often violated in biological data, which should be taken into account when constructing a mathematical model. In addition, in genome-wide association studies, the time of optimization of model parameters is a critical factor. In this paper, we propose a new SEM model, as well as a fast way to estimate its parameters.