Name Mode Size
.github 040000
R 040000
inst 040000
man 040000
src 040000
tests 040000
vignettes 040000
.Rbuildignore 100644 0 kb
.gitignore 100644 0 kb
DESCRIPTION 100644 2 kb
NAMESPACE 100644 3 kb 100644 1 kb 100644 5 kb
[![Project Status: Active – The project has reached a stable, usable state and is being actively developed.](]( [![license](]( # The `bandle` package The [**bandle**]( (Bayesian Analysis of Differential Localisation Experiments) package is an R/Bioconductor package for analysing differential localisation experiments, include storage, computation, statistics and visulisations. # Installation requirements Users will require a working version of R, currently at least version >4. It is recommended to use [RStudio]( The package can then be installed using the Bioconductor or `devtools` package. The package should take a few minutes to install on a regular desktop or laptop. The package will need to be loaded using `library(bandle)` To install the stable Bioconductor release (recommended): ```{r,} ## unless BiocManager is already installed install.packages("BiocManager") ## then BiocManager::install("bandle") ``` Unstable/development package version install: ```{r,} devtools::install_github("ococrook/bandle") ``` We do not advise you install the unstable version unless you know what you are doing as not all pre-release features may be tested or documented. ## Installation troubleshooting Please make sure you are running the latest version of R. Non-standard library dependencies may be missing on some operating systems, for example, using Linux/Ubuntu in a Docker container may requrie the installation of libxml2-dev, zlib1g-dev, libnetcdf-dev and other libraries. These are required to install `bandle` and dependent R packages. These can be installed, for example if using Linux/Ubuntu using `sudo apt install libxml2-dev` or directly from binary e.g. `sudo apt install r-cran-xml2`. # Basic ideas and concepts The [`bandle` package]( implements the [BANDLE method]( for the analysis of comparative/dynamic mass spectrometry based proteomics experiments. Data from form such experiments most commonly yield a matrix of measurements where we have proteins/peptides/peptide spectrum matches (PSMs) along the rows, and samples/fractions along the columns. In comparative/dynamic experiments where we expect re-localisation upon some stimulus to sub-cellular environment, the data analysis is more challenging The BANDLE method takes two (replicated) datasets as input and uses these data to compute the probability that a protein differentially localises upon cellular perturbation, as well quantifying the uncertainty in these estimates. To use bandle the data must be stored as a *list* of *MSnSet* instances, as implemented in the Bioconductor [`MSnbase`]( package. Please see the relevant vignettes in MSnbase for constructing these data containers. # Vignettes There are currently two vignettes that accompany the package. The first vignette [v01-getting-started]( provides an introduction to the `bandle` package and follows a short theortical example of how to perform differential localisation analysis of quantitative proteomics data using the BANDLE model (Crook et al. 2022 doi: Explanation and general recommendations of the input parameters are provided in this vignette. The second vignette [v02-workflow]( is a more comprehensive workflow which follows a real-life use case applying the BANDLE methods and workflow to the analysis of data generated from spatial proteomics of a human THP-1 monocyte system. These vignettes can be found through the Bioconductor landing page for [`bandle`](, here in the repo and also the *Articles* tab of the accompanying [web page]( Notes on run time: A small dataset can take around an hour to run; for large dataset we recommend a a compute server. The longest the analysis has taken has been a couple of hours on a single compute node. The demo take a few minutes to run. # Documentation Documentation to run the main functions can be found in the vignette or by typing `?bandle` in the console after loading the package. We recommend reading our other workflow manuscripts: Basic processing and machine learning: Bayesian analysis: The BANDLE manusript is currently on biorxiv: For manuscripts that apply bandle, see: # Contribution Contributions are welcome, please open an issue so we can discuss any contribution in advance. # Feature requests This package is actively being developed and maintained, please open Github issue if you would like to request or discuss a particular feature.