Skip to content

jeremycg/platereading

Repository files navigation

#platereading

Travis-CI Build Status

r package for reading plates off a bioanalyser to give curve parameters as an output. Plates are fit to the buchanan curve by nls (there will be bad fits) and the parameters output for further analysis.

Takes input of a directory of directories of reads. The main directory must have one directory, labelled "Plate xx", for each run. These directories must contain the .csv files given by the bioanalyser software, labelled xxx.csv, where xxx is the time in hours the reading was taken. Additionally, the root directory must contain a "strainlist.csv" which has run, column, strain and temperature columns. See the example data folder for formats

#usage

The main core of the project is now a package. To install, run:

# install.packages("devtools")
devtools::install_github("jeremycg/platereading/platereading")
library(platereading)

The batchplates.Rmd contains a step by step walkthrough of the process - it is now deprecated as most functions are in the package. The shiny app dir is also deprecated, but kept in for older users. The shiny app is now called as below. Copy and paste from below to do the main analysis:

#setwd("your data dir here"
if(file.exists("grouped.csv")){file.remove("grouped.csv")}
if(file.exists("outputfits.csv")){file.remove("outputfits.csv")}
x=namer(looper(getwd()),"strainlist.csv")
write.csv(x,file="outputfits.csv",row.names=F)
y=compileall(x,0.1)
write.csv(y,file="grouped.csv",row.names=F)

The shiny app has nice graphs and such -

plateshiny("yourdatadir")

If you already have a outputfits.csv (ie you corrected it) put it in the folder and it will not create a new one.

The shiny app allows visualisation of data based on residual cutoffs, and plots of individual fits. It is independant of the other files - they contain functions for running individual parts of the analysis. Two .csvs are generated by the shiny app - grouped.csv, a file with means and sds of each parameter, based on a cutoff residual of 0.1. The other is outputfits.csv, which has one line for each well.

##todo:

  • More intelligent handling of bad fits
  • more brands of plate readers
  • Documentation

About

r script for reading plates off a bioanalyser to QTL output

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages