Skip to main content
. Author manuscript; available in PMC: 2019 Feb 9.
Published in final edited form as: J Immunol. 2017 Aug 9:ji1700421. doi: 10.4049/jimmunol.1700421

FIGURE 3. An Elastic Net analysis identifies a cross-validated model of interrelated immune features separating patients treated with AES from controls.

FIGURE 3

(A) An elastic net (EN) algorithm extracted 20 immune features from the correlation network that differentiated the immune response to surgery between patients randomized to arginine-enriched supplements (AES) from controls (cross validated, p = 0.0001). (B) The EN features were extracted from 3 visually segregated modules. Fourteen EN features were higher and 6 features were lower in the arginine group. (C) Ten EN features projected onto module 1 (purple dots) 9 features projected onto module 2 (orange dots), and one single feature projected onto module 3 (light blue dots).