analyseColonyVectors: Maximum likelihood estimation of colony forming units per...

In JohnTownsend92/DeadOrAlive: Analysis of High-throughput Chronological Lifespan Experiments

Description

analyseColonyVectors analyses the colony patterns extracted using the extractColonyVectors function in order to estimate the number of colony forming units per droplet from the original liquid culture. This is done by performing a maximum likelihood estimation on the pattern using Brent optimisation (see optim).

Usage

analyseColonyVectors(
  colonyVectors,
  tolerance = 0.001,
  dilution = 3,
  initialDilution = 1,
  minCellsPerDroplet = 0.001,
  maxCellsPerDroplet = NULL,
  CIprob = 0.95,
  save.table = T,
  table.name = "CFUsMLE.csv",
  save.markdown = T,
  markdown.name = "CFUsMLE.html",
  save.directory = getwd()
)

Arguments

colonyVectors	An object of class colonyVectors created by extractColonyVectors containing the extracted colony patterns.
tolerance	The minimum probability of observing data at each dilution factor to be considered acceptable. If the probability of observing the data at any of the dilution factors falls below this value, the pattern will be considered anomalous. In this case, an algorithm will re-attempt the maximum likelihood estimation with excluded data points, until a suitable solution is found. Any data points which are removed are shown in the output table and markdown files.
dilution	The dilution factor used in the the experiment. Defaults to 3.
initialDilution	The initial dilution factor. Defaults to 1 (i.e. undiluted).
minCellsPerDroplet	The minimum number of viable cells per droplet to test in the optimisation. Defaults to 0.001.
maxCellsPerDroplet	The maximum number of viable cells per droplet to test in the optimisation. In the default case (NULL), a number is selected which should cover the range expected in the experiment based on the minCellsPerDroplet, dilution and length of the colony pattern.
CIprob	The probabilty of the confidence intervals to be returned. Defaults to 0.95.
save.table	Should a csv file of the results be saved. Defaults to TRUE.
table.name	Name of the csv file.
save.markdown	Should a markdown (html) document showing the results of the maximum likelihood estimation be saved. Defaults to TRUE.
markdown.name	Name of the markdown file.
save.directory	Directory to save the csv and markdown file. Defaults to the current working directory.

Value

analyseColonyVectors returns a data.frame containing the results of the maximum likelihood estimation for each sample at each time. The table contains the following columns:

Sample	The sample name.
Time	The time in days.
ColonyFormingUnitsPerDroplet	The maximum likelihood estimation of the number of viable cells per droplet.
Likelihood	The value of the maximum likelihood (i.e. the probability of observing the pattern provided in the case that the maximum likelihood estimation is correct).
LowerCI	The lower confidence interval for the estimation of viable cells per droplet.
UpperCI	The upper confidence interval for the estimation of viable cells per droplet.
GridSize	The size of the grid (i.e. the number of grid positions for each dilution factor).
ExcludedWellPositions	Any well positions which were excluded. This is as a result of an attempted optimisation not passing the tolerance threshold.
ExcludedGridPositions	Any grid positions which were excluded. This is as a result of an attempted optimisation not passing the tolerance threshold.
TotalExclusions	The total number of excluded well and grid positions.

In addition, a csv file containing this data and a markdown (html) report showing the maximum likelihood optimisation for each colony pattern are saved by default.

Examples

#Get a csv file showing the identity of each plate to be analysed
plateReferenceFile <- system.file("extdata", "plateReferenceFile.csv", package="DeadOrAlive")

#Get a csv file showing the identity of each sample on each plate
sampleReferenceFile <- system.file("extdata", "sampleReferenceFile.csv", package="DeadOrAlive")

#Get the directory of files to be analysed
dir <- system.file("extdata", "Image_Analysis", package="DeadOrAlive")

#Get the patterns of colonies from the files (from left to right across the plate)
myColonyVectors <- extractColonyVectors(dir=dir, plateReferenceFile, sampleReferenceFile)

#Perform a maximum likelihood estimation of the number of viable cells
#Note: This will save a csv and markdown file in the current working directory
## Not run: 
CFUsMLE <- analyseColonyVectors(myColonyVectors)

## End(Not run)

JohnTownsend92/DeadOrAlive documentation built on Aug. 14, 2021, 6:16 p.m.