scikit-ribo

scikit-ribo copied to clipboard

scikit-ribo

- Accurate inference and robust modelling of translation dynamics at codon resolution with Riboseq data

https://github.com/hanfang/scikit-ribo

Documentation

Read the Docs: or click me

Contact

Han Fang

Stony Brook University & Cold Spring Harbor Laboratory

Email: [email protected]

Requirement:

Environment: Python3, Linux

Recommend setting up your environment with Conda

Dependencies:

When using pip install scikit-ribo, all the following dependencies will be pulled and installed automatically.

Install

To install scikit-ribo, simply use the below command

pip install scikit-ribo

Usage

See the documentation on Read the Docs: or click me

For more information, please refer to the template shell script about details of executing the two modules.

Introduction

Scikit-ribo has two major modules: Ribosome A-site location prediction, and translation efficiency (TE) inference using a penalized generalized linear model (GLM).

A complete analysis with scikit-ribo has two major procedures:

1. data pre-processing to prepare the ORFs, codons for a genome: scikit-ribo-build.py
2. the actual model training and fitting: scikit-ribo-run.py

Inputs:

1. The alignment of Riboseq reads (bam)
2. Gene-level quantification of RNA-seq reads (from either Salmon or Kallisto)
3. A gene annotation file (gtf)
4. A reference genome for the model organism of interest (fasta)

Outpus:

1. Translation efficiency estimates for the genes
2. Translation elongation rate for 61 sense codons
3. Ribosome profile plots for each gene
4. Diagnostic plots of the models

Reference

Fang et al, "Scikit-ribo: Accurate inference and robust modelling of translation dynamics at codon resolution" (Preprint coming up)