Introduction to `immunarch`

Introduction

immunarch is an R package designed to analyse T-cell receptor (TCR) and B-cell receptor (BCR) repertoires, aimed at medical scientists and bioinformaticians. The mission of immunarch is to make immune sequencing data analysis as effortless as possible and help you focus on research instead of coding. Follow us on Twitter for news and updates.

Installation

Latest release on CRAN

In order to install immunarch execute the following command:

install.packages("immunarch")

That's it, you can start using immunarch now! See the Quick Start section below to dive into immune repertoire data analysis. If you run in any trouble with installation, take a look at the Installation Troubleshooting section.

Note: there are quite a lot of dependencies to install with the package because it installs all the widely-used packages for data analysis and visualisation. You got both the AIRR data analysis framework and the full Data Science package ecosystem with only one command, making immunarch the entry-point for single-cell & immune repertoire Data Science.

Latest release on GitHub

If the above command doesn't work for any reason, try installing immunarch directly from its repository:

install.packages("devtools") # skip this if you already installed devtools
devtools::install_github("immunomind/immunarch")

Latest pre-release on GitHub

Since releasing on CRAN is limited to one release per one-two months, you can install the latest pre-release version with bleeding edge features and optimisations directly from the code repository. In order to install the latest pre-release version, you need to execute only two commands:

install.packages("devtools") # skip this if you already installed devtools
devtools::install_github("immunomind/immunarch", ref="dev")

You can find the list of releases of immunarch here: https://github.com/immunomind/immunarch/releases

Quick start

The gist of the typical TCR or BCR data analysis workflow can be reduced to the next few lines of code.

Use immunarch data

1) Load the package and the data

library(immunarch)  # Load the package into R
data(immdata)  # Load the test dataset

2) Calculate and visualise basic statistics

repExplore(immdata$data, "lens") %>% vis()  # Visualise the length distribution of CDR3
repClonality(immdata$data, "homeo") %>% vis()  # Visualise the relative abundance of clonotypes

3) Explore and compare T-cell and B-cell repertoires

repOverlap(immdata$data) %>% vis()  # Build the heatmap of public clonotypes shared between repertoires
geneUsage(immdata$data[[1]]) %>% vis()  # Visualise the V-gene distribution for the first repertoire
repDiversity(immdata$data) %>% vis(.by = "Status", .meta = immdata$meta)  # Visualise the Chao1 diversity of repertoires, grouped by the patient status

Use your own data

library(immunarch)  # Load the package into R
immdata <- repLoad("path/to/your/data")  # Replace it with the path to your data. Immunarch automatically detects the file format.

Advanced methods

For advanced methods such as clonotype annotation, clonotype tracking, kmer analysis and public repertoire analysis see "Tutorials".

Installation troubleshooting

If you can not install devtools, check sections 1 and 2 below.

If you run in any other trouble, try the following steps:

  1. Check your R version. Run version command in the console to get your R versions. If the R version is below 3.5.0 (for example, R version 3.1.0), try updating your R version to the latest one. Note: if you try to install a package after the update and it still fails with the following message:

ERROR: dependencies ‘httr’, ‘usethis’ are not available for package ‘devtools’ * removing ‘/home/ga/R/x86_64-pc-linux-gnu-library/3.5/devtools’ Warning in install.packages : installation of package ‘devtools’ had non-zero exit status

it means that you need to re-install packages that were built under the previous R version. In the above example it would be packages httr and usethis. In order to re-install a package you need to execute the command install.packages("package_name"), where package_name is the name of the package to update. To find packages that need to be re-installed after updating R, you need to look for installation messages like this in the installation process:

ERROR: package ‘usethis’ was installed by an R version with different internals; it needs to be reinstalled for use with this R version

  1. Check if your packages are outdated and update them. In RStudio you can run the "Update" button on top of the package list in the "Package" window. In R console you can run the old.packages() command to view a list of outdated packages. The following messages indicate that an update is required:

Error: package ‘dtplyr’ 0.0.3 was found, but >= 1.0.0 is required by ‘immunarch’ Execution halted ERROR: lazy loading failed for package ‘immunarch’

byte-compile and prepare package for lazy loading Error in loadNamespace(j <- i[[1L]], c(lib.loc, .libPaths()), versionCheck = vI[[j]]) : namespace 'ggalluvial' 0.9.1 is being loaded, but >= 0.10.0 is required Calls: <Anonymous> ... namespaceImportFrom -> asNamespace -> loadNamespace Execution halted

  1. For Mac users. Make sure to install XCode from App Store first and command line developers tools second by executing the following command in Terminal: xcode-select –install

  2. For Mac users. If you have issues like old packages can't be updated, or error messages such as ld: warning: directory not found for option or ld: library not found for -lgfortran, this link will help you to fix the issue.

  3. For Mac Mojave (1.14) users. If you run into the following error:

``` /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/include/c++/v1/math.h:301:15: fatal error: 'math.h' file not found

include_next

          ^~~~~~~~

```

Open Terminal, execute the following command and try again to install immunarch:

sudo installer -pkg /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg -target /

  1. For Linux users. If you have issues with the igraph library or have Fortran errors such as:

** testing if installed package can be loaded from temporary location Error: package or namespace load failed for 'igraph' in dyn.load(file, DLLpath = DLLpath, ...): unable to load shared object '/usr/local/lib/R/site-library/00LOCK-igraph/00new/igraph/libs/igraph.so': libgfortran.so.4: cannot open shared object file: No such file or directory

See this link for help.

  1. For Linux users. If you have issues with the rgl package:

configure: error: missing required header GL/gl.h ERROR: configuration failed for package ‘rgl’

Install "mesa-common-dev" via OS terminal by executing the following command:

apt-get install mesa-common-dev

Check this link for more information and other possible workarounds.

  1. If you have error messages with rlang in them such as:

Error: .onLoad failed in loadNamespace() for 'vctrs', details: call: env_bind_impl(.env, list3(...), "env_bind()", bind = TRUE) error: object 'rlang_env_bind_list' not found

Remove the rlang package and install it again. This error is often happens after updating R to a newer version, while rlang not being properly updated.

  1. If you have error messages like the following (note the (converted from warning) part):

** byte-compile and prepare package for lazy loading Error: (converted from warning) package 'ggplot2' was built under R version 3.6.1 Execution halted ERROR: lazy loading failed for package 'immunarch'

Execute the following command in R and try again to install the package:

Sys.setenv(R_REMOTES_NO_ERRORS_FROM_WARNINGS="true")

  1. For Windows users. If you have issues with the package installation, or if you want to change the folder for R packages, feel free to check this forum post.

  2. For Windows users. Make sure to install Rtools. Before installation close RStudio, install Rtools and re-open it afterwards. To check if Rtools installed correctly, run the devtools::find_rtools() command (after installing the devtools package). If you have an error, check this link for help.

  3. If you can not install dependencies for immunarch, please try manual installation of all dependencies by executing the following command in R console:

install.packages(c("rematch", "prettyunits", "forcats", "cellranger", "progress", "zip", "backports", "ellipsis", "zeallot", "SparseM", "MatrixModels", "sp", "haven", "curl", "readxl", "openxlsx", "minqa", "nloptr", "RcppEigen", "utf8", "vctrs", "carData", "pbkrtest", "quantreg", "maptools", "rio", "lme4", "labeling", "munsell", "cli", "fansi", "pillar", "viridis", "car", "ellipse", "flashClust", "leaps", "scatterplot3d", "modeltools", "DEoptimR", "digest", "gtable", "lazyeval", "rlang", "scales", "tibble", "viridisLite", "withr", "assertthat", "glue", "magrittr", "pkgconfig", "R6", "tidyselect", "BH", "plogr", "purrr", "ggsci", "cowplot", "ggsignif", "polynom", "fastcluster", "plyr", "abind", "dendextend", "FactoMineR", "mclust", "flexmix", "prabclus", "diptest", "robustbase", "kernlab", "GlobalOptions", "shape", "colorspace", "stringi", "hms", "clipr", "crayon", "httpuv", "mime", "jsonlite", "xtable", "htmltools", "sourcetools", "later", "promises", "gridBase", "RColorBrewer", "yaml", "ggplot2", "dplyr", "dtplyr", "dbplyr", "data.table", "gridExtra", "ggpubr", "pheatma3", "ggrepel", "reshape2", "DBI", "factoextra", "fpc", "circlize", "tidyr", "Rtsne", "readr", "readxl", "shiny", "shinythemes", "treemap", "igraph", "airr", "ggseqlogo", "UpSetR", "stringr", "ggalluvial", "Rcpp"))

  1. If you encounter the following error while running the devtools::install_local function:

``` In normalizePath(path.expand(path), winslash, mustWork) : path[1]="path/to/your/folder/with/immunarch.tar.gz":

In file.copy(x$path, bundle, recursive = TRUE) : problem copying No such file or directory ```

Check your path to the downloaded package archive file. It should not be "path/to/your/folder/with/immunarch.tar.gz", but a path on your PC to the downloaded file, e.g., "C:/Users/UserName/Downloads/immunarch.tar.gz" or "/Users/UserName/Downloads/immunarch.tar.gz".

  1. If troubles still persist, let us know via GitHub (preferably) or support@immunomind.io (in case of private data).


Try the immunarch package in your browser

Any scripts or data that you put into this service are public.

immunarch documentation built on July 7, 2021, 9:08 a.m.