/paper-5utr-design

Code for "Optimizing 5’UTRs for mRNA-delivered gene editing using deep learning"

Primary LanguageJupyter NotebookBSD 3-Clause "New" or "Revised" LicenseBSD-3-Clause

Optimizing 5’UTRs for mRNA-delivered gene editing using deep learning

This repository contains data analysis and sequence design code in the publication "Optimizing 5’UTRs for mRNA-delivered gene editing using deep learning", including code to generate all Figure panels.

Contents

  • analysis_cell_type: Code for analysis of cell type data in Figure 1 and related Supplementary Figures.
  • analysis_random_end: Code for analysis of the random-end MPRA libraries from Figure 3 and related Supplementary Figures.
  • megatal_5utr_design: Notebooks and scripts for 5'UTR design, used in megaTAL gene editing assays.
  • megatal_gene_editing_analysis: Code for analysis of megaTAL gene editing results
  • models_sample2019: Deep learning models from Sample, et al. Nat. Biotech 2019. must be downloaded and placed here for the code in this repo to work. See README file inside.
  • polysome_profiling_data: Polysome profiling data acquired in this study must be placed here for the code in this repo to work. See README file inside.
  • polysome_profiling_sample2019: Polysome profiling data acquired in Sample, et al. Nat. Biotech 2019. must be downloaded and placed here for the code in this repo to work. See README file inside.

Requirements

All code was written in Python 3. With the exception of the sequence design code (megatal_5utr_design), the following package versions were used:

  • matplotlib 3.5.1
  • numpy 1.22.1
  • pandas 1.4.3
  • scikit-learn 1.0.2
  • scipy 1.7.3
  • seaborn 0.12.2
  • tensorflow 2.7.0

For sequence design, the following are required:

All standard packages can be installed with pip or conda. For isolearn, Fast SeqProp, and DEN, follow the instructions in the respective repos.

Usage

Most of the analysis code is in Jupyter notebooks, each of which can be opened and run to reproduce a specific analysis or design task.