wukevin / rnagps

Interpretable model of sub-cellular RNA localization.

RNA-GPS

Interpretable model for predicting high-resolution RNA sub cellular localization to the following localizations:

• ER Membrane
• Nuclear lamina
• Mito matrix
• Cytosol
• Nucleolus
• Nucleus
• Nuclear pore
• Outer mito membrane

This model is trained on APEX-seq data, which measures RNA localization human HEK293T cells.

SARS-CoV-2 Analysis

Since viruses reproduce by hijacking human cellular machinery, we can also use this model to generate hypotheses surrounding localization of SARS-CoV-2 RNA transcripts. See analyses in the covid19 directory for additional information, as well as relevent works below.

Installation

Download the codebase via git clone and use the following command to create the rnagps conda environment.

conda env create -f environment.yml

Activate the environment with conda activate rnagps.

Usage

After installing the conda environment, you can make predictions by using the script under bin/predict_localization.py as follows:

python bin/predict_localization.py <5'UTR sequence> <CDS equence> <3'UTR sequence>

Relevant works

• Wu, K.E., Parker, K.R., Fazal, F.M., Chang, H., and Zou, J. (2020). RNA-GPS predicts high-resolution RNA subcellular localization and highlights the role of splicing. RNA. link
• Wu, K.E., Fazal, F.M., Parker, K.R., Zou, J., and Chang, H.Y. (2020). RNA-GPS Predicts SARS-CoV-2 RNA Residency to Host Mitochondria and Nucleolus. Cell Systems. link