Supplementary Materials

Supplementary Materials for:

Drug Synergy Screen and Network Modeling in Dedifferentiated Liposarcoma Identifies CDK4 and IGF1R as Synergistic Drug Targets

Martin L. Miller,* Evan J. Molinelli,* Jayasree S. Nair, Tahir Sheikh, Rita Samy, Xiaohong Jing, Qin He, Anil Korkut, Aimee M. Crago, Samuel Singer,* Gary K. Schwartz,* Chris Sander*

*Corresponding author. E-mail: liposarcoma_combo@cbio.mskcc.org will reach the principal authors

This PDF file includes:

  • Fig. S1. WDLS/DDLS tumors and the two cell lines used in this study have complex DNA copy number alterations.
  • Fig. S2. EC50 determination with a cell metabolic activity assay shows good agreement between biological replicates.
  • Fig. S3. Effects of PD0332991 and NVP-AEW541 in DDLS8817 and LPS141 cells.
  • Fig. S4. CDK4 inhibition causes G1 cell cycle arrest in LPS141 cells.
  • Fig. S5. Phosphorylation of AKT in DDLS8817 cells is consistently suppressed for at least 24 hours after PI3K inhibition.
  • Fig. S6. Examples of determining the appropriate drug concentration for RPPA-based proteomic profiling in DDLS8817 cells.
  • Fig. S7. Western blot analysis of antibodies used in the RPPA assay shows no apparent cross-reactivity.
  • Fig. S8. mRNA expression of nodes in the network falls in the top half of all genes in both cell lines used.
  • Fig. S9. DNA copy number analysis of network genes shows copy gains in multiple genes and amplification of CDK4.
  • Fig. S10. Schematic illustration of the computational analysis on a fictional four-node system.
  • Fig. S11. Network models are not overfitted to prior knowledge interactions.
  • Fig. S12. BP improves the performance of prior knowledge interactions alone.
  • Fig. S13. BP inference is minimally sensitive to drug specificity.
  • Fig. S14. Liposarcoma-specific network models are predictive of cell response to drugs.
  • Fig. S15. Many of the drug combinations with the strongest predicted synergy scores are categorized in accordance with experiments.
  • Fig. S16. Combined inhibition of the CDK4 and IGF1R nodes is predicted to be synergistic by the network models.
  • Fig. S17. Combination treatment enhances repression of mTOR signaling compared to single-drug treatment.
  • Fig. S18. Inhibition of EGFR and CDK4 has synergistic effects on cell metabolic activity, and effects are enhanced in a triple perturbation adding an IGF1R inhibitor.
  • Fig. S19. Combining CDK4 inhibition with MEK or ERK inhibition does not result in synergistic effects on cell viability based on cell metabolic activity.
  • Fig. S20. Many of the top 100 models predict synergistic effects of combined CDK4 and IGF1R inhibition.
  • Table S1. Drugs used in the synergy screen (cell viability) and the proteomic screen (RPPA).
  • Table S2. Dose-response measurements of single- and paired-drug perturbations with the resazurin assay.
  • Table S3. Combination index scores.
  • Table S4. Drugs used in follow-up experiments.
  • Table S5. Interactions in the prior knowledge network.
  • Table S6. Bayesian-derived models have many interactions in common with BP-derived models.
  • References (58–96)

[Download PDF]

Technical Details

Format: Adobe Acrobat PDF

Size: 4.23 MB


Citation: M. L. Miller, E. J. Molinelli, J. S. Nair, T. Sheikh, R. Samy, X. Jing, Q. He, A. Korkut, A. M. Crago, S. Singer, G. K. Schwartz, C. Sander, Drug Synergy Screen and Network Modeling in Dedifferentiated Liposarcoma Identifies CDK4 and IGF1R as Synergistic Drug Targets.. Sci. Signal. 6, ra85 (2013).

© 2013 American Association for the Advancement of Science