biogrowth (version 1.0.8)

fit_serial_dilution: Serial-fold dilution method

Description

Model fitting by the serial-fold dilution method

Usage

fit_serial_dilution(
  TTD_data,
  start,
  dil_factor = 2,
  mode = "intercept",
  logN_det = NULL,
  logN_dil0 = NULL,
  max_dil = NULL
)

Arguments

TTD_data

a tibble (or data.frame) with the TTD observed for different dilutions. It must have two columns: TTD (the TTD) and dil the number of serial dilutions.

start

named numeric vector of initial guesses for the model parameters

dil_factor

dilution factor. By default, 2

mode

one of "intercept" (serial dilution method with a generic intercept; default) or "lambda" (able to estimate also the value of the lag phase duration)

logN_det

log10 microbial concentration at the detection OD (only for mode = "lambda")

logN_dil0

log10 microbial concentration at wells where dilution = 0 (only for mode = "lambda")

max_dil

maximum number of dilutions to include. By default, NULL (no limit)

Examples

## We can use the example data set

data("example_od")

## We first need to estimate the TTDs

library(tidyverse)

my_TTDs <- get_TTDs(example_od, target_OD = 0.2, codified = TRUE)
my_data <- filter(my_TTDs, condition == "S/6,5/35/R1")

## Fitting using the "intercept" mode

guess <- c(a = 0, mu = .1)  # we need initial guesses for the model parameters

my_fit <- fit_serial_dilution(my_data, start = guess)

## The class returned implements common S3 methods

my_fit
summary(my_fit)
plot(my_fit)

## The fitting can define a maximum number of dilutions

my_fit <- fit_serial_dilution(my_data, start = guess, max_dil = 4)
plot(my_fit)

## Fitting using the "lambda" mode

logNdet <- 7.5  # this mode requires the microbial concentration at the detection OD
logN_dil0 <- 4  # and the concentration at the well with dilution 0 
guess <- c(lambda = 0, mu = .1)  # the guess must be defined now on lambda instead of a

my_fit2 <- fit_serial_dilution(my_data, 
                               start = guess,
                               mode = "lambda", 
                               logN_det = logNdet,
                               logN_dil0 = logN_dil0)

## The instance implements the same S3 methods as before

my_fit2
summary(my_fit2)
plot(my_fit2)