Fatty acid metabolic profiles and biomarker discovery for type 2 diabetes mellitus using graphical index of separation combined with principal component analysis and partial least squares-discriminant analysis

► A strategy of plasma fatty acid profiles was employed to investigate T2DM. ► GIOS-PCA was applied to select biomarkers of T2DM patients and healthy control. ► PLS-DA was used to build discriminate model of selected variables for T2DM patients. ► C22:6, C20:5, 16:0 and C20:2 were biomarkers for assisting FPG in T2DM diagnosis. ► T2DM patients with different FPG values possess no significant metabolic difference.