predict_RBPs is a framework for predicting RNA-binding protein (RBPs) on Protein Occupancy Profile-Sequencing (POP-seq) data, where POP-seq is protocol for capturing nearly all RBP-RNA interactions, regardless of their identities and it doesn’t require crosslinking.
python 3.6.10
numpy 1.18.1
pandas 1.0.1
sklearn 0.22.2.post1
tensorflow 2.0.0
keras 2.3.1 (using Tensorflow backend)
In order to run predict_RBP, the user has to do the following:
1- Ensure that Encode data file that that identifies the RBPs on the CLIP-seq data a specific cell line is in the same path where main.py file exists.
2- Ensure that POP-seq data file for the same cell line on which the RBPs will be predicted/identified is in the same path where main.py file exists.
3- Rrn generate_bench_mark.py to generate the benchmark dataset for POP-seq data on which popseq data is labeled by RBPs based on the intersection of genomic locations on CLIP-seq and POP-seq data
4- Run the following python command:
python main.py
The default Machine Learning model used in predict_RPBs framework is the Random Forest (RF) with RNA sequence embedding using Word2Vec technique.