the analysis script for PrismNet paper: Predicting dynamic cellular protein-RNA interactions using deep learning and in vivo RNA structure
The scripts are used to calculate the structure change distance score of transcipt regions and and search the structurally variable sites between different cell. We provided the result of structurally variable sites between each pair of cell lines.
We provided some examples of PrismNet prediction result, including the RBPs binding probability of sliding window (101nt) in transcripts of different cell lines.
The scripts are used to calculate the correlation between PrismNet predicted RBP binding and RNA half-life, translation efficiency and splicing.
The scripts are used for the integrative motif construction, motif clustering, motif significant analysis and motif scanning. We provided the result of integrative motif of PrismNet model.
The scripts are used to search high attention region (HAR).
The scripts are used for riboSNitch finding and the analysis of relationship between riboSNitches and human disease. We provided the result of riboSNitch between each pair of cell lines.