Overview

HTTPS SSH

DDR a method to predict drug target interactions using multiple similarities

Dependencies:

  • Python 2.7
  • numpy
  • Scikitlearn

Input format and files:

  • DDR expects all network files to in form of adjacency list file.
  • For relation files DDR expect a tuple of drug and target in each line
  • For similarity files DDR expects a tuple of drug (target) and drug (target) and their similarity

Usage

usage: DDR.py [-h] --interaction R_FILE --DSimilarity D_SIM_FILE --TSimilarity T_SIM_FILE --outfile OUT_FILE [--no_of_splits NO_OF_SPLITS] [--K K] [--K_SNF K_SNF] [--T_SNF T_SNF] [--N NO_OF_TREES] [--s SPLIT]

DDR a method to predict drug target interactions

optional arguments: -h, --help show this help message and exit --no_of_splits NO_OF_SPLITS Number of parts to split unkown interactions. Default: 10 --K K Number of nearest neighbors for drugs and targets neigborhood. Default: 5 --K_SNF K_SNF Number of neighbors similarity fusion. Default: 3 --T_SNF T_SNF Number of iteration for similarity fusion. Default: 10 --N NO_OF_TREES Number trees for random forest. Default: 100 --s SPLIT Split critera for random forest trees. Default: gini

required named arguments: --interaction R_FILE Name of the file containg drug target interaction tuples --DSimilarity D_SIM_FILE Name of the file containg drug similarties file names --TSimilarity T_SIM_FILE Name of the file containg target similarties file names --outfile OUT_FILE Output file to write predictions

Contact

  • Vladimir Bajic (vladimir-dot-bajic-at-kaust-dot-edu-dot-sa)
  • Rawan Olayan (rawan-dot-olayan-at-kaust-dot-edu-dot-sa))