BPP - branch point prediction

BPP is a method that can identify branch point based on only the intron sequence.

How BPP works?

BPP predicts the branch point sequence by integrating the degenerative motif of BPS and PPT characteristics. Specifically, BPP uses a mixture model to infer the BPS motif and a set of weighted octanucleotides to estimate the contribution of the 65kDa subunit of U2AF (U2AF65).

Detailed information on BPSP and citation

A paper describing BPP is under review.

Dependent libraries or software

• Python3.5

Quickstart

Usage: BP_PPT.py -b -p -i -r -h

Required:

• -b, --pwm file STR The file including PWM of BPS
• -p, --ppt file STR The file including the PPT score
• -i, --FASTA file STR The file including the fasta sequence
• -r, --report nu INT The reported sites; default=1; 0: print all positions
• -h, --help

Example:

• BP_PPT.py -b demo/pwmBP_human.txt -p demo/scPPT_human.txt -i demo/example.fa

Format of the output file:

• id: ID of the intron
• bps: the branch point sequence
• bp_pos: the position of the branch point relative to 3'SS upstream
• sc_bps: the score of the BPS
• sc_ppt: the score of the PPT
• sc: the score of the BPS and PPT
• zsc_bps: the z-score of the BPS
• zsc_ppt: the z-score of the PPT
• zsc: the z-score of the BPS and PPT