Skip to content

Prepare Sailfish and Salmon output for downstream analysis

License

Notifications You must be signed in to change notification settings

COMBINE-lab/wasabi

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

40 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Important note for Windows users: If you are running Wasabi under R or R studio on Windows, there is a known issue (bug, I would say) that prevents normal operation. Salmon and Sailfish output important extra information like bootstrap samples and the fragment length distribution to a subdirectory of the quantification directory named aux. Windows, unfortunately, forbids aux as a folder name. If you are lucky enough to be reading this before running Sailfish/Salmon, you can pass an alternative folder name to the --auxDir command line option to avoid this issue. Otherwise, we suggest the following work-around. First, rename the aux subdirectory of each Sailfish/Salmon quantification directory to be a folder name that is not forbidden under Windows e.g. aux2. Then, open the cmd_info.json file in the quantification directory, and add the following line:

"auxDir" : "aux2",

Now, things should work normally under windows. Salmon versions (>=0.7.0) use aux_info as the default folder to avoid this issue. Future releases of Sailfish (>0.10.1) will use a different default name for the auxiliary directory to avoid this issue as well.

What is wasabi?

Join the chat at https://gitter.im/COMBINE-lab/wasabi

Wasabi allows you to easily prepare Sailfish and Salmon output for downstream analysis.
Currently, its main purpose it to prepare output for downstream analysis with sleuth.

How to use wasabi

Installation

First, you need to install the wasabi package. There are two main ways to accomplish this:

Installation with devtools

With devtools, it's easy to install wasabi:

source("http://bioconductor.org/biocLite.R")
biocLite("devtools")    # only if devtools not yet installed
biocLite("COMBINE-lab/wasabi")

Installation with bioconda

Alternatively, you can use the conda package manager, along with the bioconda channel to install wasabi:

conda create -n wasabi r-wasabi

Loading and using wasabi

Once wasabi is installed, you can load the library with:

library(wasabi)

Now that wasabi is installed, we can use it to convert the Sailfish / Salmon output into sleuth-compatible format. Imagine we have some samples (Sailfish quant directories) sitting in a data directory:

data/samp1   data/samp2   data/samp3   data/samp4

First, we create a simple vector containing these directories (the > below is an R prompt):

> sfdirs <- file.path("data", c("samp1", "samp2", "samp3"))

Now, we simply run the prepare_fish_for_sleuth function:

> prepare_fish_for_sleuth(sfidrs)

The function will write some status messages to the console and, when it's done, each directory will now contain and abundance.h5 file in a sleuth-compatible format. From this point forward, you can simply run sleuth normally.