RiboFlow

RiboFlow is a Nextflow based pipeline for processing ribosome profiling data. As output, it generates ribo files that can be analyzed using RiboR or RiboPy. RiboFlow belongs to a software ecosystem desgined to work with ribosome profiling data.

Contents

• Installation
• Test Run
• Output
• RiboFlow on Your Data
• A Note on References
• Advanced Features
• Frequently Asked Questions
• Release Notes

Installation

Requirements

• Nextflow
• Docker (Optional)
• Conda (Optional)

First, follow the instructions in Nextflow website and install Nextflow.

The easiest way of using RiboFLow is using Docker. If using Docker is not an option, you can install the dependencies using Conda and run RiboFlow without Docker.

Docker Option

Install Docker. Here is a tutorial for Ubuntu.

All remaining dependencies come in the Docker image ceniklab/riboflow. This image is automatically pulled by RiboFlow when run with Docker (see test runs below).

Conda Option

This option has been tested on Linux systems only.

Install Conda.

All other dependencies can be installed using the environment file, environment.yaml, in this repository.

git clone https://github.com/ribosomeprofiling/riboflow.git
conda env create -f riboflow/environment.yaml

The above command will create a conda environment called ribo and install dependencies in it. To start using RiboFlow, you need to activate the ribo environment.

conda activate ribo

Test Run

For fresh installations, before running RiboFlow on actual data, it is recommended to do a test run.

Run Using Docker

# Clone this repository in a new folder and change your working directory to the RiboFlow folder.
mkdir rf_test_run && cd rf_test_run
git clone https://github.com/ribosomeprofiling/riboflow.git
cd riboflow

# Obtain a copy of the sample data in the working directory.
git clone https://github.com/ribosomeprofiling/rf_sample_data.git
nextflow RiboFlow.groovy -params-file project.yaml -profile docker_local

Note that we provided the argument -profile docker_local to Nextflow to indicate that RiboFlow will be run via Docker containers. In other words, the steps of RiboFlow will be executed inside Docker containers by Nextflow. Hence, no locally installed software (other than Java and Nextflow) is needed by RiboFlow.

Run Using Conda Environment

In Conda option, the steps of RiboFlow are run locally. So, we need to install the dependencies first. This can easily be done via conda. The default profile directs RiboFlow to run locally, so we can simply skip the -profile argument. Also note that the conda environment has to be activated before running RiboFlow.

Before running the commands below, make sure that you have created the conda environment, called ribo, using the instructions above.

# List the environments to make sure that ribo environment exists
conda env list

# Activate the ribo environment
conda activate ribo

# Get RiboFlow repository
mkdir rf_test_run && cd rf_test_run
git clone https://github.com/ribosomeprofiling/riboflow.git
cd riboflow

# Obtain a copy of the sample data in the working directory.
git clone https://github.com/ribosomeprofiling/rf_sample_data.git

# Finally run RiboFlow
nextflow RiboFlow.groovy -params-file project.yaml

Output

Pipeline run may take several minutes. When finished, the resulting files are in the ./output folder.

Mapping statistics are compiled in a csv file called stats.csv

ls output/stats/stats.csv

Ribosome occupancy data is in a single ribo file called all.ribo.

ls output/ribo/all.ribo

You can use RiboR or RiboPy to work with ribo files.

RiboFlow on Your Data

For running RiboFlow on actual data, files must be organized and a parameters file must be prepared. You can examine the sample run above to see an example.

1. Organize your data. The following files are required for RiboFlow

   • Ribosome profiling sequencing data: in gzipped fastq files
   • Transcriptome Reference: Bowtie2 index files
   • Filter Reference: Bowtie2 index files (typically for rRNA sequences)
   • Annotation: A bed file defining CDS, UTR5 and UTR3 regions.
   • Transcript Lengths: A two column tsv file containing transcript lengths

2. Prepare a custom project.yaml file. You can use the sample file project.yaml, provided in this repository, as template.

3. In project.yaml, provide RiboFlow parameters such as clip_arguments, alignment arguments etc. You can simply modify the arguments in the sample file project.yaml in this repository.

4. You can adjust the hardware and computing environment settings in Nextflow configuration file(s). For Docker option, see configs/docker_local.config. If you are not using Docker, see configs/local.config.

5. RNA-Seq data is optional for RiboFlow. So, if you do NOT have RNA-Seq data, in the project file, set

do_rnaseq: false

If you have RNA-Seq data to be paired with ribosome profiling data, see the Advanced Features below.

6. Metadata is optional for RiboFlow. If you do NOT have metadata, in the project file, set

do_metadata: false

If you have metadata, see Advanced Features below.

7. Run RiboFlow using the new parameters file project.yaml.

Using Docker:

nextflow RiboFlow.groovy -params-file project.yaml -profile docker_local

Without Docker:

nextflow RiboFlow.groovy -params-file project.yaml

A Note on References

RiboFlow is designed to work with transcriptomic references. RiboFlow does NOT work with genomic references. The users need to provide a transcriptome reference and annotation to run this software. There is a curated set of RiboFlow references, that users can download and use, in this GitHub repository

Advanced Features

RNA-Seq Data

If you have RNA-Seq data that you want to pair with ribosome profiling experiments, provide the paths of the RNA-Seq (gzipped) fastq files in the configuration file in input -> metadata. See the file project.yaml in this repository for an example. Note that the names in defining RNA-Seq files must match the names in definig ribosome profiling data. Also turn set the do_rnaseq flag to true, in the project file:

do_rnaseq: true

Transcript abundance data will be stored in the output ribo file.

Metadata

If you have metadata files for the ribosome profiling experiments, provide the paths of the metadata files (in yaml format) in the configuration file in input -> metadata. See the file project.yaml in this repository for an example. Note that the names in defining metadata files must match the names in definig ribosome profiling data. Also turn set the metadata flag to true, in the project file:

do_metadata: true

Metadata will be stored in the output ribo file.

Citing

RiboFlow, RiboR and RiboPy: an ecosystem for analyzing ribosome profiling data at read length resolution, H. Ozadam, M. Geng, C. Cenik Bioinformatics 36 (9), 2929-2931

@article{ozadam2020riboflow,
  title={RiboFlow, RiboR and RiboPy: an ecosystem for analyzing ribosome profiling data at read length resolution},
  author={Ozadam, Hakan and Geng, Michael and Cenik, Can},
  journal={Bioinformatics},
  volume={36},
  number={9},
  pages={2929--2931},
  year={2020},
  publisher={Oxford University Press}
}

Frequently Asked Questions

Release Notes