ribofy: ORF detection using RiboSeq data

Ribofy is a fast and simple python-based tool for detection of phased p-sites across open-reading-frames (ORFs)

Installation

• pip (soon)

pip install ribofy

• from source

git clone https://github.com/ncrnalab/ribofy.git
cd ribofy
python setup.py install

Running ribofy

First, all ORFs are assembled from an annotation file (preferably gencode GTF) and the corresponding genome fasta (should not take more than 5-10 minutes). This is only required once per genome/annotation:

ribofy orfs --gtf <path/to/gtf> --fa <path/to/fasta>

The genome fasta-file must be indexed prior to ORF assembly:

samtools faidx <path/to/fasta>

Currently, ribofy is compatible with STAR, kallisto and salmon mapped reads. Recommended mapping commands:

• STAR

STAR --genomeDir <path/to/STAR_index> --outSAMtype BAM SortedByCoordinate 
--readFilesIn <path/to/fastq_files> --readFilesCommand zcat --outFileNamePrefix <sample_prefix>. 

• salmon

salmon quant -i </path/to/salmon_index -l A </path/to/fastq_files> --gcBias --validateMappings {additional_params} --writeMappings=</path/to/output.bam> -o </path/to/output>

• kallisto

kallisto quant -i </path/kallisto_index> --bias -o </path/to/output> --single --pseudobam --fr-stranded -l 30 -s 2 </path/to/fastq_files>

Note that for kallisto and salmon, genome indexing should be performed with reduced k-mer value to allow mapping of <30nt ribosome-protected fragments.

Before running ribofy, bam-files should be sorted and indexed:

samtools sort </path/to/bamfile> > </path/to/sorted/bam_file>
samtools index </path/to/sorted/bam_file>

Then, run ribofy:

ribofy detect --orfs <path/to/orf/assembly> --bams </path/to/bamfiles> --prefix <prefix>

Under the hood

1. Ribofy infers the p-site offsets for read-lengths between 25 and 35 (although this can be customized) and outputs the <prefix>.offset.txt

2. Then, for each ORF, ribofy counts the p-sites and evaluates the statistical enrichment of in-frame p-sites. This outputs the <prefix>.phasing.txt

3. Finally, Ribofy collects the individual ORFs into ORF-groups (collapsing overlapping and correlating ORFs), preserving only the highest expressed ORF (based on overall coverage), performs ORF-type specific FDR corrections and outputs the final <prefix>.results.txt

Citation

in preparation

Contact

Thomas Hansen ([email protected])