Predicting PPI Based on Quantum-inspired Neural Networks

EasyChair Preprint 1108

Abstract

Studying how and when PPI (Protein-Protein Interaction) happens in cells is very important for catching molecules mechanism during life. Unfortunately, present classical computers technology and process speed for massive PPI data is far from meeting demand. It is a general trend of researching that utilizing quantum computation methods, bioinformatics knowledge and machine learning ability to effectively investigate existing massive data for discovering and verifying new PPI. In this paper, we utilize the quantum ant colony optimization algorithm (QAC) to optimize the quantum-inspired neural network QNN’s parameters and propose a novel quantum neurocomputing scheme QAC-QNN which can prevent the solution sinking into local optimization. The simulation experiments prove that the novel scheme is feasible. For predicting PPI of proteins couple, in contrast with SVM and ANN, QAC-QNN achieves better prediction effects.

Keyphrases: Ant Colony Optimization, Bioinformatics, PPI, neural network, quantum computation

@booklet{EasyChair:1108,
  author    = {Li-Ping Yang and Cheng Zhang and Li Qin},
  title     = {Predicting PPI Based on Quantum-inspired Neural Networks},
  howpublished = {EasyChair Preprint 1108},
  year      = {EasyChair, 2019}}