Improve workflow to extend the structure of an existing PBPK Model (e.g. add Tumor Organ)

[msevestre]

The current workflow to extend a PK-Sim structure in MoBi (for example in order to add an Tumor organ to a model) could be simplified greatly.

At the moment, the workflow consists of several steps, most of which could be automatized by MoBi:

1. Save a representative organ as .pkml file (for example Heart)
2. Import the saved organ as container and name it Tumor
3. Rename all internal neighborhoods between sub-containers of imported organs
4. Modify tag of new organ & sub-containers (e.g. Rename Heart tag to Tumor)
5. Create neighborhoods for new organ (i.e, "wire" it to the blood flow)
6. Update sum parameter Weight (tissue) in Organism
7. Update sum parameter Volume(plasma) in Organism
8. Update sum parameter Blood Flow Rate in Lung
9. Define the partition coefficient (int/water)

I believe that steps 3, 4, 6, 7, 8, 9 should be performed automatically:

• Step 3: Rename of internal neighborhoods should happen automatically. If Heart was renamed to Tumor, then heart_bc_heart_pls neighborhood should also be renamed to tumor_bc_tumor_pls etc..
• Step 4: Renaming of Tags in subcontainers should also happen automatically
• Steps 6,7,8 are very similar. The definition of parameters Weight (tissue) in Organism, Volume(plasma) in Organism, Blood Flow Rate in Lung should be changed to used a generic SumFormula instead of listing the list of parameters explicitly. This would have to be implemented in PK-Sim
• Step 9: The parameter is added automatically by the calculation method. Instead, this parameter should be defined all the time in the spatial structure. If it is not used, the overhead is minimal. This would have to be implemented in PK-Sim

What do you think?

[PavelBal]

This is a crucial feature for extending models. So far, the workflow is error-prone and takes much time and effort. Implementing your suggestions would help a lot.

[Yuri05]

Even more steps are required in case of protein model. All of them can be also automated via sum formulas

10. Volume of Endosome in EndogenousIgG/Endosome: Add endosomal tumor volume
11. Volume of Interstitial in EndogenousIgG/Interstitial: Add interstitial tumor volume
12. Volume of Plasma in EndogenousIgG/Plasma: Add plasma tumor volume
13. Lymph flow rate in EndogenousIgG: Add tumor lymph flow rate
14. Fluid recirculation flow rate in EndogenousIgG: Add tumor fluid recirculation flow rate
15. Surface area (plasma/interstitial) in the Neighborhood EndogenousIgG_int_EndogenousIgG_pls: Add respective surface area from tumor.

Another additional steps are not related to sum formulas:

16. Add new equation for the volume of endosomal space in tumor Note also that additional neighborhoods are needed with the protein model: a) Tumor endosomal space <-> EndosomalClearance compartment b) Tumor interstitial space <-> VenousBlood plasma
17. The property is present has to be adapted in the Molecule Start Values if the tumor is added manually.

[msevestre]

@Yuri05 I really think that we should try to tackle step 9 Define the partition coefficient (int/water) asap. This is something that is confusing for everybody and just does not make sense.

[StephanSchaller]

Great!

[wilbertdew]

Hi All,

Thanks for this excellent thread! I have been looking for a while why my tumor interstitial concentrations were different compared to the organ which was duplicated (muscle) while only the volume had changed. Turned out to be a combination of one of the points from Yuri and a new one:

16. Add new equation for the volume of endosomal space in tumor
17. Add new equation for the volume of vascular endothelium in tumor (i.e., refer to the tumor neighborhood rather then the original organ)

In addition if you change the volume of the tumor compared to the duplicated organ, the following is also relevant for the correct calculation of the isogravimetric flow if I am not mistaken:

19. Make sure the "Volume" parameter value of the tumor is identical to the "Organ volume mouse" parameter value of the tumor

If anyone has additional points/different experience or if there is a more recent thread, I would be happy to hear.

P.S. I cannot easily find this thread either via the issues or discussions search functionality, maybe possible to move this issue to the discussions?

Best,

Wilbert

[Yuri05]

Make sure the "Volume" parameter value of the tumor is identical to the "Organ volume mouse" parameter value of the tumor

Hmm, not sure...

I cannot easily find this thread either via the issues or discussions search functionality

Show all open issues in all OSP repositories (then you can add search keywords to narrow the search results) https://github.com/issues?q=is%3Aopen+is%3Aissue+archived%3Afalse+user%3AOpen-Systems-Pharmacology+sort%3Aupdated-desc

Show all discussions in all OSP repositories: https://github.com/discussions?discussions_q=user%3AOpen-Systems-Pharmacology

maybe possible to move this issue to the discussions?

Absolutely :)