Construct a symmetric matrix from its lower triangle.

Description

Construct a symmetric matrix from its lower triangle.

Usage

LTmat(LT, .names = attr(LT, ".names"), .colnames = .names, .rownames = .names)

Arguments

Value

A symmetric matrix.

Examples

LTmat(1:6)

Coerce a pkprofile to a data.frame

Description

Coerce a pkprofile to a data.frame

Usage

## S3 method for class 'pkprofile'
as.data.frame(x, ..., col.names = c("time", "conc"), .state = FALSE)

Arguments

Value

A data.frame with columns time and conc (or the names specified in col.names). If .state == TRUE, then the complete state is appended (as a matrix column).

Construct a block-diagonal matrix.

Description

Construct a block-diagonal matrix.

Usage

blockdiag(...)

Arguments

Value

A block-diagonal matrix.

Examples

blockdiag(matrix(1, 2, 2), 2, matrix(3, 4, 4))

Convert from standard deviation and correlation matrix to covariance matrix.

Description

Convert from standard deviation and correlation matrix to covariance matrix.

Usage

cor2cov(cor, sd)

Arguments

Value

A covariance matrix.

Examples

cor2cov(matrix(c(1, 0.5, 0.5, 1), 2, 2), 0.1)

Get the doses from a PK profile.

Description

Get the doses from a PK profile.

Usage

dose.frame(x)

Arguments

Value

A data.frame containing the realized doses, one per row. The data.frame has all the columns described in pkprofile, except addl, since all additional doses have been expanded to individual rows. It also has a conc column with the simulated concentration at the time of the dose.

See Also

pkprofile

Examples

t.obs <- seq(0, 6*24, 0.5)
y <- pkprofile(t.obs, cl=0.5, vc=11, ka=1.3,
    dose=list(t.dose=c(0, 24*2 + 14), amt=c(100, 50), addl=c(4, 0), ii=24))
dose.frame(y)

Get the final state or time of a PK profile.

Description

Get the final state or time of a PK profile.

Usage

finalstate(x)

finaltime(x)

Arguments

Value

A numeric vector containing the state of each compartment at the final observation time (finalstate), or the final observation time itself (finaltime).

See Also

• pkprofile for generating a PK profile.

• pkprofile.pkprofile for appending to an existing PK profile.

Examples

# Administer a dose at time 0 and a second dose using the final state
# from the first dose (at 12h) as the initial state for the second dose.
t.obs <- seq(0, 12, 0.1)
y <- pkprofile(t.obs, cl=0.25, vc=5, ka=1, dose=list(t.dose=0, amt=1))
finalstate(y)
y2 <- pkprofile(t.obs, cl=0.25, vc=5, ka=1, dose=list(t.dose=0, amt=1), initstate=finalstate(y))
plot(y, xlim=c(0, 24), ylim=c(0, max(y2)), col="blue")  # First dose
lines(t.obs+12, y2, col="red")                          # Second dose

# Add a vertical line to show where the first profile ends.
abline(v=finaltime(y), col="gray75", lty=2)

Generate individual random effects from a multivariate normal distribution.

Description

Generate individual random effects from a multivariate normal distribution.

Usage

generateETA(n, omegaLT, omega = LTmat(omegaLT), eta.names = colnames(omega))

Arguments

Value

An n \times p matrix, where each row contains the vector of random effects for one individual (p is the size of the covariance matrix).

See Also

LTmat blockdiag

Examples

omegaLT <- c(0.123, 0.045, 0.678)
generateETA(10, omegaLT)

Half-lives of a linear PK system.

Description

Half-lives of a linear PK system.

Usage

halflife(x)

Arguments

Value

A numeric vector containing the half-lives for the different phases of the system. The number of phases generally equal the number of compartments, plus one for the absorption phase if the system has first order absorption (i.e. if ka is specified). The values are returned sorted in ascending order, so the first corresponds to the alpha phase, the second beta, the third gamma, and so on. The absorption half-life, if present, comes last (it can also be identified by comparing it to the value of log(2)/ka).

Examples

y <- pkprofile(0, cl=0.25, vc=5, ka=1.1)
halflife(y)
log(2)/1.1

y <- pkprofile(0, cl=0.25, vc=5, ka=0.01)  # Flip-flop kinetics
halflife(y)
log(2)/0.01

# Three-compartment model
y <- pkprofile(0, cl=2, vc=10, q=c(0.5, 0.3), vp=c(30, 40))
halflife(y)

# The terminal half-life can be used to obtain the terminal slope of the
# concentration-time curve on the semi-log scale:
t.obs <- seq(0, 36, 0.1)
y <- pkprofile(t.obs, cl=0.25, vc=5, ka=1, dose=list(t.dose=0, amt=1))
plot(log2(y))
abline(-2.247927, -1/halflife(y)[1], col=adjustcolor("blue", 0.2), lwd=12)

Runs the interactive shiny app.

Description

Runs the interactive shiny app.

Usage

linpkApp(...)

Arguments

Value

Called for its side effects.

Note

The app requires the following packages:

• 'shiny'

• 'shinyjs'

• 'shinyAce'

• 'dygraphs'

Make they are installed or the app won't work.

Examples

## Not run: 
linpkApp()

## End(Not run)

Generate a concentration-time profile.

Description

This function generates concentration-time profiles from a linear pharmacokinetic (PK) system, possibly with first-order absorption or zero-order infusion, possibly with one or more peripheral compartments, and possibly under steady-state conditions. Single or multiple doses may be specified.

Usage

pkprofile(...)

## Default S3 method:
pkprofile(
  t.obs = seq(0, 24, 0.1),
  cl = 1,
  vc = 5,
  q = numeric(0),
  vp = numeric(0),
  ka = 0,
  dose = list(t.dose = 0, amt = 1, rate = 0, dur = 0, ii = 24, addl = 0, ss = 0, cmt = 0,
    lag = 0, f = 1),
  sc = vc,
  initstate = NULL,
  ...
)

## S3 method for class 'matrix'
pkprofile(
  A,
  t.obs = seq(0, 24, 0.1),
  dose = list(t.dose = 0, amt = 1, rate = 0, dur = 0, ii = 24, addl = 0, ss = 0, cmt = 0,
    lag = 0, f = 1),
  defdose = 1,
  sc = 1,
  initstate = NULL,
  ...
)

Arguments

Value

An object of class "pkprofile", which is simply a numeric vector of concentration values with some attributes attached to it. This object has its own methods for print, plot, lines and points.

Methods (by class)

• pkprofile(default): Default method.

• pkprofile(matrix): Matrix method.

Warning

Pay attention to the default arguments. They are there for convenience, but may lead to undesired results if one is not careful.

See Also

• halflife

• secondary

• print.pkprofile

• plot.pkprofile

• lines.pkprofile

• points.pkprofile

Examples

# Default values, a bolus injection
y <- pkprofile()
plot(y)

t.obs <- seq(0, 24, 0.1)
dur <- 1
amt <- 1
ka <- 1
cl <- 0.25
vc <- 5
q <- 2.5
vp <- 10

# One-compartment model with first-order absorption, single dose
y <- pkprofile(t.obs, cl=cl, vc=vc, ka=ka, dose=list(amt=amt))
plot(y)

# Two-compartment model with first-order absorption, single dose
y <- pkprofile(t.obs, cl=cl, vc=vc, vp=vp, q=q, ka=ka, dose=list(amt=amt))
plot(y)

# One-compartment model with zero-order infusion, single dose
y <- pkprofile(t.obs, cl=cl, vc=vc, dose=list(dur=dur, amt=amt))
plot(y)

# Two-compartment model with zero-order infusion, single dose
y <- pkprofile(t.obs, cl=cl, vc=vc, vp=vp, q=q, dose=list(dur=dur, amt=amt))
plot(y)

# Two-compartment model with bolus injection, single dose
y <- pkprofile(t.obs, cl=cl, vc=vc, vp=vp, q=q, dose=list(amt=amt))
plot(y)

# Two-compartment model with bolus injection into the peripheral compartment, single dose
y <- pkprofile(t.obs, cl=cl, vc=vc, vp=vp, q=q, dose=list(amt=amt, cmt=2))
plot(y)

# Two-compartment model with zero-order infusion into the peripheral compartment, single dose
y <- pkprofile(t.obs, cl=cl, vc=vc, vp=vp, q=q, dose=list(amt=amt, cmt=2, dur=dur))
plot(y)

t.obs <- seq(0, 24*6, 1)

# One-compartment model with first-order absorption, multiple doses
y <- pkprofile(t.obs, cl=cl, vc=vc, ka=ka, dose=list(t.dose=seq(0, 24*5, 12), amt=amt))
plot(y)

# One-compartment model with first-order absorption, multiple doses specified by addl and ii
y <- pkprofile(t.obs, cl=cl, vc=vc, ka=ka, dose=list(t.dose=0, amt=amt, addl=9, ii=12))
plot(y, type="b")
points(y, col="blue")

# One-compartment model with first-order absorption, multiple doses under steady-state conditions
yss <- pkprofile(t.obs, cl=cl, vc=vc, ka=ka, dose=list(t.dose=0, amt=amt, addl=9, ii=12, ss=1))
lines(yss, col="red")
points(yss, col="green")

# One-compartment model with zero-order infusion, multiple doses specified by addl and ii
y <- pkprofile(t.obs, cl=cl, vc=vc, dose=list(dur=dur, amt=amt, addl=9, ii=12))
plot(y, log="y")

# One-compartment model with zero-order infusion, multiple doses  under steady-state conditions
yss <- pkprofile(t.obs, cl=cl, vc=vc, dose=list(dur=dur, amt=amt, addl=9, ii=12, ss=1))
lines(yss, col="red")

Printing and plotting methods for class pkprofile.

Description

Printing and plotting methods for class pkprofile.

Usage

## S3 method for class 'pkprofile'
print(x, ...)

## S3 method for class 'pkprofile'
plot(x, y, ...)

## S3 method for class 'pkprofile'
lines(x, y, ...)

## S3 method for class 'pkprofile'
points(x, y, ...)

Arguments

Continue an existing concentration-time profile.

Description

This method can be used to append to an existing PK profile, for instance to simulate a PK profile with parameters that change over time. Each time the parameters change, a new call to this method is used to advance the system with the new parameter values.

Usage

## S3 method for class 'pkprofile'
pkprofile(obj, t.obs = finaltime(obj) + seq(0, 24, 0.1), ..., append = TRUE)

Arguments

Value

An object of class "pkprofile".

Warning

The new parameters take effect at the time when the previous profile ends. If the previous profile ends before the new sampling starts, the new parameters will be used to advance the system to the start of the new sampling.

Any ongoing zero-order infusion at the end of the previous profile is dropped. The remaining infusion amount will NOT be carried forward.

See Also

pkprofile

Examples

t.obs <- seq(0, 24, 0.1)
amt <- 1
ka <- 1
cl <- 0.25
vc <- 5

# One-compartment model with first-order absorption
# First dose at time 0
y <- pkprofile(t.obs, cl=cl, vc=vc, ka=ka, dose=list(t.dose=0, amt=amt))

# Second dose at 24h with a lower clearance
y <- pkprofile(y, t.obs+24, cl=0.5*cl, vc=vc, ka=ka, dose=list(t.dose=24, amt=amt))

# Third dose at 48h with a higher clearance
y <- pkprofile(y, t.obs+48, cl=2*cl, vc=vc, ka=ka, dose=list(t.dose=48, amt=amt))
plot(y)

Derive secondary PK parameters.

Description

Derive secondary PK parameters.

Usage

secondary(x, From = NULL, To = NULL, include.dose.times = T)

Arguments

Value

A data.frame with one row for each time interval and with the following columns:

From

The time of the start of the interval. Can differ from the specified start time because it always corresponds to an actual data point.

To

The time of the end of the interval. Can differ from the specified end time because it always corresponds to an actual data point.

N

The number of distinct data points in the interval used to derive AUC, Cmax, etc.

Ctrough

Concentration at the time of dose (i.e. just prior to the dose). Only present if the start of the interval corresponds to a dose time.

Cmin

Minimum concentration over the interval.

Tmin

Time of the minimum concentration over the interval.

Cmax

Maximum concentration over the interval.

Tmax

Time of the maximum concentration over the interval.

Cave

Average concentration over the interval (calculated by the trapezoid rule).

AUC

Area under the concentration-time curve over the interval (calculated by the trapezoid rule).

Examples

t.obs <- seq(0, 24*4, 0.1)
y <- pkprofile(t.obs, cl=0.25, vc=5, ka=1, dose=list(t.dose=0, amt=1, addl=6, ii=12))
secondary(y)
secondary(y, 0, 48)
secondary(y, 0, Inf)
sum(secondary(y)$AUC)  # Same as above
plot(y)
with(secondary(y), points(Tmax, Cmax, pch=19, col="blue"))
with(secondary(y), points(Tmin, Cmin, pch=19, col="red"))
with(secondary(y), points(From, Ctrough, pch=19, col="green"))
with(secondary(y), points(From + 6, Cave, pch=19, col="purple", cex=2))