Fig 1: Personalized predictive cancer models were created for MM cell line MM.1S (ATCC CRL-2974) and MM cell line U266B1 (ATCC TIB-196). Single cell and multi-cell computational models became ‘personalized’ when cell line-specific genomic data were annotated into MM computational models and validated when predicted responses were compared to chemokine, cytokine, and cell-associated biomarker responses from the same cell lines grown alone or with DC in multi-cell cultures. (A–D) There were differences in the predicted and observed CD47, FASL, and PD-L1 responses and the IL6, IL-10 TGFB1, and VEGFA responses between MM.1S and U266B1. (E) When ‘personalized,’ U266B1 was predicted to have higher concentrations of IL-10, IL6, VEGFA, and PD-L1 than MM.1S and these predictions were validated by measuring the IL-10, IL6, VEGFA, and PD-L1 concentrations from MM.1S and U266B1 grown in single cell cultures. When the MM.1S vs. U266B1 responses were compared, the U266B1 > MM.1S matched 75% (3/4). (F) In multi-cell computational models with DC, U266B1 and MM.1S were predicted to inhibit DC CD80, CD86, IL2, IFNG, and IL12B responses. The percent change with respect to control was greatest with U266B1 > MM.1S. (G) Production of DC markers CD80, CD86, IL2, IFNG, and IL12B were lower when DC were cultured in multi-cell cultures with U266B1 or MM.1S: U266B1 attenuated DC marker production more than MM.1S and these responses matched 100% (6/6).