Ensemble of Template-Free and Template-Based Classifiers for Protein Secondary Structure Prediction.

Gabriel Bianchin De Oliveira,Helio Pedrini,Zanoni Dias

doi:10.3390/ijms222111449

Gabriel Bianchin De Oliveira, Helio Pedrini + Show 1 more

Open Access

https://doi.org/10.3390/ijms222111449

Copy DOI

Abstract

Protein secondary structures are important in many biological processes and applications. Due to advances in sequencing methods, there are many proteins sequenced, but fewer proteins with secondary structures defined by laboratory methods. With the development of computer technology, computational methods have (started to) become the most important methodologies for predicting secondary structures. We evaluated two different approaches to this problem—driven by the recent results obtained by computational methods in this task—(i) template-free classifiers, based on machine learning techniques; and (ii) template-based classifiers, based on searching tools. Both approaches are formed by different sub-classifiers—six for template-free and two for template-based, each with a specific view of the protein. Our results show that these ensembles improve the results of each approach individually.

Highlights

Proteins are present in many biological processes in the cells of living beings, playing different functions, such as transport, growth, and maintenance of the body
Template-Free Classifiers In this subsection, we present the experimental evaluation of the template-free classifiers
We present our six template-free classifiers—bidirectional recurrent neural network, random forest, inception-v4 blocks, inception recurrent network, Bidirectional Encoder Representations from Transformers (BERT), and convolutional neural network

Summary

Introduction

Proteins are present in many biological processes in the cells of living beings, playing different functions, such as transport, growth, and maintenance of the body. They are formed by a sequence of amino acids, which consist of the protein’s primary structure [1]. Amino acids interact physically and chemically with each other, forming three-dimensional structures. The local three-dimensional structure that each amino acid participates in is called the secondary structure, whereas three-dimensional structure that the protein forms is called the tertiary structure [2]. The most common method in the literature is to first understand the secondary structure and predict the tertiary structure

Objectives

Methods

Results

Discussion

Conclusion