search

Open-Systems-Pharmacology / Forum

Discussion forum for the Open Systems Pharmacology Project
65 stars 19 forks source link

Strange values for partition coefficient #353

Closed MahmAbdelwahab closed 5 years ago

MahmAbdelwahab commented 5 years ago

Hello,

I'm new to using Pk-sim, just practicing with a basic PBPK model for clofazimine, which his highly lipophilic drug that accumulate mainly in adipose and fatty tissues mainly metabolized by liver (Cl : 0.26 l/h/kg) given the published physicochemical characteristics of the drug,

Effective M.w: 429 (drug bank) Fu : 0.01 (published study) Logp : 7.66 (drug bank) Water solubility : 0.233 mg/l (drug bank) PKa is 8.6 When Pk-sim calculate (B/P ratio and Pc's), it gives me strange values, for example the B/p value of > 1 (2615.2) which I know is not correct, but if I log10 the value I get reasonable values for B/p ratio and Pc's.

What I'm missing or doing wrong ?

Thanks

Mahmoud

StephanSchaller commented 5 years ago

Some initial thoughts:

In our experience, DrugBank LogP is often too high by some margin. LogD (@pH 7.4) or LogMA is more likely to be the right input for PK-Sim. Also an fu of 0.01 is probably measured with high uncertainty. Increasing fu should also lower B:P.

What Partitioning Calculation Method did you use? For a highly lipophilic drug, I would recommend Rodgers and Rowland.

If you have data for B:P try first setting R&R as the calculation method, then decrease lipophilicity (Calculated LogP of ~7 often compare to LogD ~5) and maybe increase fu and see where you end up.

Hope this helps, Stephan

msevestre commented 5 years ago

Please reopen if needed