Skip to main content
Fig. 3 | BMC Bioinformatics

Fig. 3

From: Multivariate pattern analysis: a method and software to reveal, quantify, and visualize predictive association patterns in multicollinear data

Fig. 3

The model structure displaying the relations between groups of variables. HOMA-IR, which is a proxy for insulin resistance, is predicted from two groups of explanatory variables, i.e., three measures of adiposity and a physical activity descriptor providing the number of counts at 23 intensity intervals derived from an accelerometric sensor. Age and sex are confounders, and a profile of 26 lipoprotein measures acts as a mediator between HOMA-IR and adiposity and physical activity. Arrows imply the directions of the relations

Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us