The Algorithm Of Detecting The Critical Transition Point Of Complex Diseases

In today's society,many complex diseases seriously threaten people's quality of life and even their safety.These complex diseases have the phenomenon of sudden deterioration.From systems biology point of view,the disease system with time evolution has critical point.Apparently,to detect the critical point of complex diseases is of great significance for disease prevention and control work.In this paper,from the viewpoint of systems biology and bioinformatics,using the theory and method of computational biology based on highthroughput biological molecular data and dynamic network,we study how to detect the critical point of disease.Under the condition of four different samples we develope the following four kinds of algorithms:(1)Based on molecular expression,a unsupervised learning algorithm is proposed.We state the ideal of the algorithm,that is,let the switching points of stationary markov process corresponding to the critical point of disease.We give the steps of the algorithm and we apply it to detect the critical point of breast cancer.(2)Based on molecular expression and regulation network,we give another unsupervised learning algorithm.We propose methods to build the molecular correlation network,specificity network and the sequence of diversity network.Also the network model is construction,and the simulation test is made.This algorithms is applied to detect the critical point of acute lung injury caused by phosgene inhalation.(3)We use torgue development to reduce the data noise.We give the methods to proposed the dimentions of the state variables,expand the torgue to k-orders,deal with the closed torque,simulate the system with random disturbance.The algorithm is applied to detect the critical point of liver cancer,and the function of the signal genes is analyzed.(4)We propose a local network entropy algorithm which including building the local network with single sample,formula of network entropy,calculation of comprehensive network entropy and system simulation.Finally,the algorithm is applied to determine the critical points of lung cancer and infection virus infection.