sORF

Python scripts to extract sequence-derived features of small ORFs and perform RF-based validation and prediction

featureCombineTrain.py

Get all features of train dataset

train5fold.py

Grid search based on five-fold validation in Prokaryotic species so that I can validate the best parameters (tree number, depth, sample number for splitting) of RF-based model for the prediction

modelPredict.py

Training and cross-species prediction and obtain the performance

featureCombine.py

Get all the features of testing datasets

main.py

Grid search based on five-fold validation in eukaryotic species so that I can validate the best parameter of RF model for the prediction

tryAll.py

Cross-species prediction (One to one)

gridSearchMixed.py

Merge all kind of features of all species and perform cross-species prediction (Four to one)

featurePackage.py

Including some functions to calculate the sequence-derived features, which is called by the above scripts