Skip to content

Repository files navigation

Translational Efficiency Prediction

This project predicts the translational efficiency (TE) of genes from RNA sequence features using machine learning. It models log-transformed TE values under eIF3d and eIF4e depletion conditions with Ridge regression, ElasticNet, and LightGBM models. Features are extracted from transcript sequences and merged into a single dataset for training.

How to Run

  1. Unzip TE_eIF_depletion.zip
  2. Run the .py files in main to extract sequence features from the raw data in TE_eIF_depletion.csv (this takes up hundreds of MBs)
  3. (optional) Run check.py as a sanity check on all extracted .csv files
  4. Run merge.py to get the entire input dataframe
  5. Run your desired model in its directory

We did it this way because we each worked on feature extractions separately:

Feature Author
Length features Jannes
Nucleotide frequency Anuschka
Codon and aa frequency Oliver
K-mer frequency Jannes
Wobble position Anuschka
One-hot encoding of nt around start codon Jannes
Counts of 20 relevant dicodons according to Gamble et al Jannes
Secondary structure values Oliver

Instructions for LGBM

Running LGBM.py will create a folder with directories for each of the defined feature subsets along with a general summary of the model results.

Inside the feature subset directories are two folders with the model results for the two depletion conditions (eIF3d and eIF4e). They contain .csv files with the search results across the folds (currently 5), feature importance scores, and predicted vs observed values on the test set. A model summary is also given as a .json file.

Currently the feature subsets are the ones described in the report.

About

Predicting change in translational efficiency under initiation factor depletion. Dataset by MDC

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages