Time after last dose as a time-varying covariate

How-to
1

Hello all,

I would like to use time after last dose as a time-varying variable in differential equations. Is there any convenient way to do this? Though I saw a similar thread discussing tad in the dataset, I am posting here as my focus is time after dose in the @model. If not, would adding a column of time of last doing in the dataset and subtracting in the @model be the best way?

2

First, there are some restrictions on covariates names (as columns in a DataFrame):

The following names are restricted form being used as covariate names: id , amt , time , evid , cmt , rate , duration , ss , ii , route , and tad . These names are used for columns in the DataFrame output form various Pumas objects, and therefore the names would clash when adding covariate information as columns.

Source: Defining NLME models in Pumas Β· Pumas

Finally, you can use the tad function inside a Pumas @model, see this MWE:

model = @model begin
    @param begin
        tvcl ∈ RealDomain(lower = 0)
        tvv ∈ RealDomain(lower = 0)
        Ω ∈ PDiagDomain(2)
        Οƒ_prop ∈ RealDomain(lower = 0)
    end
    @random begin
        Ξ· ~ MvNormal(Ξ©)
    end
    @pre begin
        CL = tvcl * exp(Ξ·[1])
        Vc = tvv * exp(Ξ·[2])
    end
    @dynamics Central1

    @derived begin
        cp := @. 1000 * (Central / Vc)
        dv ~ @. Normal(cp, sqrt(cp^2 * Οƒ_prop))
    end
    @observed begin
        mytad = tad(t, events)
    end
end

ev = DosageRegimen(100; addl = 2, ii=1)

pops = map(i -> Subject(id = i, events = ev), 1:2)

sims = simobs(model, pops, init_params(model); obstimes = 0.5:3.5)

DataFrame(sims)

output:

julia> DataFrame(sims)
14Γ—17 DataFrame
 Row β”‚ id       time     dv             mytad      evid    amt        cmt      rate       duration   ss       ii         route       Ξ·_1        Ξ·_2       Central           CL        Vc
     β”‚ String?  Float64  Float64?       Float64?   Int64?  Float64?   Symbol?  Float64?   Float64?   Int8?    Float64?   NCA.Route?  Float64?   Float64?  Float64?          Float64?  Float64?
─────┼─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
   1 β”‚ 1            0.0  missing        missing         1      100.0  Central        0.0        0.0        0        0.0  NullRoute    0.434505  -1.44498  missing           1.5442    0.235751
   2 β”‚ 1            0.5    36690.0            0.5       0  missing    missing  missing    missing    missing  missing    missing      0.434505  -1.44498        3.78144     1.5442    0.235751
   3 β”‚ 1            1.0  missing        missing         1      100.0  Central        0.0        0.0        0        0.0  NullRoute    0.434505  -1.44498  missing           1.5442    0.235751
   4 β”‚ 1            1.5     1433.44           0.5       0  missing    missing  missing    missing    missing  missing    missing      0.434505  -1.44498        3.78685     1.5442    0.235751
   5 β”‚ 1            2.0  missing        missing         1      100.0  Central        0.0        0.0        0        0.0  NullRoute    0.434505  -1.44498  missing           1.5442    0.235751
   6 β”‚ 1            2.5    19090.6            0.5       0  missing    missing  missing    missing    missing  missing    missing      0.434505  -1.44498        3.78686     1.5442    0.235751
   7 β”‚ 1            3.5       -8.84295        1.5       0  missing    missing  missing    missing    missing  missing    missing      0.434505  -1.44498        0.00541495  1.5442    0.235751
   8 β”‚ 2            0.0  missing        missing         1      100.0  Central        0.0        0.0        0        0.0  NullRoute   -0.615916   1.89947  missing           0.540146  6.68237
   9 β”‚ 2            0.5    30734.0            0.5       0  missing    missing  missing    missing    missing  missing    missing     -0.615916   1.89947       96.039       0.540146  6.68237
  10 β”‚ 2            1.0  missing        missing         1      100.0  Central        0.0        0.0        0        0.0  NullRoute   -0.615916   1.89947  missing           0.540146  6.68237
  11 β”‚ 2            1.5    55271.4            0.5       0  missing    missing  missing    missing    missing  missing    missing     -0.615916   1.89947      184.621       0.540146  6.68237
  12 β”‚ 2            2.0  missing        missing         1      100.0  Central        0.0        0.0        0        0.0  NullRoute   -0.615916   1.89947  missing           0.540146  6.68237
  13 β”‚ 2            2.5    48414.2            0.5       0  missing    missing  missing    missing    missing  missing    missing     -0.615916   1.89947      266.324       0.540146  6.68237
  14 β”‚ 2            3.5   127547.0            1.5       0  missing    missing  missing    missing    missing  missing    missing     -0.615916   1.89947      245.643       0.540146  6.68237

Check the function tad here: Pumas Docstrings Β· Pumas

1 Like