dyusuf / RCAS_alpha

RCAS project

Make a standalone RNA Centric Annotation System that provides intuitive reports and publication ready graphics.

RCAS takes input peak intervals in BED foramt from clip-seq data and automatically generates distributions of annotation features, detected motifs, GO-term enrichment, pathway enrichment and genomic coverage.

The workflow is modularized so user can switch on/off optional steps including motif detection, GO-term enrichment, pathway enrichment and genomic coverage. By default these optional steps are switched off and RCAS only generates distributions of annotation features including biotypes and genomic features.

Installing and building from source

RCAS uses autotools to provide a standard configuration script and installation rules following the GNU coding standards. To build from a simple repository checkout you need to first bootstrap the configuration script.

git clone https://github.com/BIMSBbioinfo/RCAS.git
cd RCAS
autoreconf -vif
./configure --prefix=/where/you/want/rcas
make install

When building from a release tarball this bootstrap step is not required:

tar xf rcas-0.0.1.tar.gz
cd rcas
./configure --prefix=/where/you/want/rcas
make install

If your system provides both Python 2.x and Python 3.x and defaults to Python 3.x, you may need to pass PYTHON=/path/to/python2 to the configure script to ensure that the Python 2.x interpreter is used, for example:

./configure PYTHON=/bin/python2 --prefix=/opt/rcas
sudo make install

After the first official release we will also provide a Guix package definition for RCAS so that you can install the pipeline with a single command such as:

guix package -i rcas

Hacking on RCAS

For development purposes we also provide a package expression for GNU Guix. To enter a development shell in which all dependent tools are available run this:

guix environment -l package.scm

In this shell you can configure and install RCAS as usual, but all tools and R packages will be handled by Guix. There is no need to manually install the tools.

RCAS dependencies

We recommend to install dependencies with a package manager using one of the package recipes we provide (see "Hacking on RCAS" above). If you really must install the dependencies manually please see the following list for pointers:

Tools

• snakemake
• R
• fastaFromBed from bedtools
• MEME-chip
• pandoc (>= 1.12.3)

R packages

• rmarkdown
• rtracklayer
• data.table
• biomaRt
• org.Hs.eg.db
• org.Ce.eg.db
• org.Mm.eg.db
• org.Dm.eg.db
• topGO
• DT
• plotly
• dplyr
• genomation
• GenomicFeatures

Data files

• gff track e.g. wget -qO- ftp://ftp.sanger.ac.uk/pub/gencode/Gencode_human/release_19/gencode.v19.annotation.gff3.gz | gunzip > gencode.v19.annotation.gff3
• genome reference e.g. this hg19 reference
• c2.cp.v5.0.entrez.gmt

RCAS workflow is built with snakemake:

Note that files containing placeholders that are substituted at configuration time end on .in. Successful configuration generates the target files without the .in extension. In the lists below we omit .in for clarity.

• The pipeline wrapper and entry point:

  • src/RCAS

• All snakemake rules end on .snakefile.

  • src/RCAS.snakefile
  • src/RCAS.anot.snakefile
  • src/RCAS.motif.snakefile
  • src/RCAS.GOrich.snakefile
  • src/RCAS.PATHrich.snakefile

• Custom scripts that are called at runtime:

  • src/parse_anot.py
  • src/top_motifs.py
  • src/rcas.GO.R
  • src/rcas.msigdb.R
  • src/rcas.Rmd
  • src/generate_report.sh

• Plain data files are located in src/base:

  • Homo_sapiens-U2T.meme
  • Mus_musculus-U2T.meme
  • Caenorhabditis_elegans-U2T.meme
  • Drosophila_melanogaster-U2T.meme
  • c2.cp.v5.0.entrez.gmt
  • c2.cp.v5.0.entrez.dm3.gmt
  • c2.cp.v5.0.entrez.mm9.gmt
  • c2.cp.v5.0.entrez.ce6.gmt
  • custom.css
  • header.html
  • img folder

Test case

Aim: generate analysis report for intervals in the target BED files.

Besides the default annotation step, pathway enrichment is enabled via the --run_PATHrich flag.

1. Install as per the instructions above, e.g.

   autoreconf -vif
   ./configure --prefix=/opt/rcas
   make clean
   sudo make install
   

2. cd /opt/rcas

3. target files in ./test: xaa.bed, xab.bed

4. check help message: ./bin/RCAS -h

5. sample command:

   ./bin/RCAS --genome /path/to/hg19.fa                   \
              --gff3 /path/to/gencode.v19.annotation.gff3 \
              --run_PATHrich                              \
              ./test/xaa.bed ./test/xab.bed
   

6. output: xaa.rcas.html, xab.rcas.html