Skip to content

Latest commit

 

History

History
40 lines (31 loc) · 3.37 KB

changelog.md

File metadata and controls

40 lines (31 loc) · 3.37 KB

Previous modifications were made based on QSP model by Jafarnejad et al. (AAPS 2019):

  • Added MDSC parameter file and MDSC module.
  • Added default parameter file with baseline parameter for TNBC
  • Added hill functions for MDSC and ENT in Cancer killing (T cell module) and Cancer cell growth (cancer module); and added hill function parameters in simbio_init.m.
  • Updated pk parameter file and pk module to add entinostat.
  • Updated schedule_dosing.m to add entinostat dose regimen.

Modifications made based on the previous version published on Front. Bioeng. Biotechnol. 2020 (doi.org/10.3389/fbioe.2020.00141):

  • Added CD4 T helper cell module, and modified nomenclature of Teff and Treg parameters
  • Modified T cell module, including naive T cell dynamics, cytokine secretions, cancer cell killing, and T cell trafficking parameters
  • Added nab-paclitaxel PK/PD module
  • Modified H_APC calculation to avoid overestimation of Treg density at the beginning of the simulation
  • Changed tumor growth dynamic to Gompertzian model for dynamic maximal tumor capacity
  • Added dynamic PD-L1 expression that is upregulated by IFN-gamma secreted by T helper cells
  • Modified Treg expansion mechanism that is induced by arginase I and TGF-beta

Modifications made since the published version in J Immunother Cancer. 2021 (DOI: 10.1136/jitc-2020-002100):

  • Added a new macrophage module and a submodule for checkpoint interactions between cancer cell and macrophage that inhibit phagocytosis (macrophage_module.m, phagocytosis_module.m)
  • Modified MDSC module to recalibrated cellular recruitment and cytokine secretion rates to match clinical measurements from TNBC tumors (MDSC_module.m, parameters_TNBC.m)
  • Modified parameter nomenclature to be more consistent among the modules (parameters_TNBC.m, PSA_param_in_TNBC.m)
  • Added estimated distribution of cytokine secretion rates and macrophage/MDSC recruitment rates for virtual patient generation (PSA_param_in_TNBC.m)
  • Added postprocessing steps to calculate macrophage, MDSC, M1/M2 macrophage ratio, etc. (calculate_ratio.m, sprint_data.m)
  • Added T cell death upon antigen clearance (Teff_module.m, Treg_module.m, Th_module.m)

Modifications made since the published version in iScience 2022 (DOI: 10.1016/j.isci.2022.104702):

  • Added parameter files for NSCLC and scripts for virtual patient generation (new methodology) & clinical trial simulation of durvalumab in NSCLC.
  • Added a script for PK parameter fitting using compressed latent parameterization & updated durvalumab PK parameters & dosing schedule.
  • Updated a model assumption so that IFN-gamma is now secreted by both CD8 and CD4 T cells when activated.
  • Modified nomenclature for some parameters & reorganized structure for some modules (to maintain relative independence & customizability of the modules, delete redundant model elements, and allow selection of Gompertzian/logistic growth dynamic in cancer_module.m).
  • Added a script for sensitivity analysis using Morris screening method to account for non-monotonicity of the model.

Modifications made since the published version in NPJ Precision Oncology (DOI: 10.1038/s41698-023-00405-9):

  • Added PK parameter for anti-CD47 antibodies.
  • Incoporated dynamics of anti-CD47 antibodies to PK module and phagocytosis module.
  • Modified simbio_PSA.m so that the parameter values can be changed at the beginning of in silico clinical to simulate treatment effect.