Method for efficiently predicting Gram-negative bacteria type III and type IV effector proteins

Abstract

The invention discloses a method for efficiently predicting Gram-negative bacteria type III and type IV effector proteins. The method is a deep neural network framework (CHR) method, and the method comprises the following steps: (1) constructing an integrated deep neural network framework; (2) selecting a data set; (3) using the two-dimensional features and the three-dimensional features as input features of the network; (4) learning a prediction module on the data set by using the built integrated deep neural network framework; (5) setting model parameters; (6) inputting a to-be-detected protein sequence into the model to obtain a protein prediction result. The method not only has a certain reference effect on related research in the future, but also has important significance on knowing related biological functions of gram-negative bacteria.

Description

technical field [0001] The invention relates to a method for predicting effector proteins of Gram-negative bacteria, in particular to a method for efficiently predicting Gram-negative bacteria type III and type IV effector proteins Background technique [0002] With the rapid development of omics technology, microbiology research has also entered a new stage of development. Through microbiology, we can observe the microorganisms and their composition in different natural environments, and clearly understand the role of these microorganisms in many aspects such as human health, environmental restoration, agricultural production, and marine ecology. The macroscopic functions of microorganisms are often the result of the joint action of a complex group of different microorganisms. Therefore, in order to understand the macroscopic functions of microorganisms more clearly, it is necessary to study the physiological activities of microorganisms from the microscopic level. There a...

AI Technical Summary

Problems solved by technology

The premise of in-depth research on T3SEs and T4SEs is to be able to predict them quickly and accurately, although many methods have been used to predict T3SEs and T4SEs, (it is recommended to list the main prediction methods in the prior art and the literature recorded, and list the shortcomings Relative to) but some of these methods are time-consuming and costly, some have low prediction accuracy for effector proteins, and some can only be applied to the prediction of one effector protein. Therefore, how to develop a lightweight Universal tools, enabling efficient prediction of these two effector proteins, remain an important biological challenge

Images