DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes

Abstract Deep learning models (DLMs) have gained importance in predicting, detecting, translating, and classifying a diversity of inputs. In bioinformatics, DLMs have been used to predict protein structures, transcription factor-binding sites, and promoters. In this work, we propose a hybrid model t...

Full description

Saved in:
Bibliographic Details
Main Authors: Leonardo Ledesma-Dominguez, Erik Carbajal-Degante, Gabriel Moreno-Hagelsieb, Ernesto Pérez-Rueda
Format: Article
Language:English
Published: Nature Portfolio 2024-04-01
Series:Scientific Reports
Online Access:https://doi.org/10.1038/s41598-024-59487-5
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846165346779660288
author Leonardo Ledesma-Dominguez
Erik Carbajal-Degante
Gabriel Moreno-Hagelsieb
Ernesto Pérez-Rueda
author_facet Leonardo Ledesma-Dominguez
Erik Carbajal-Degante
Gabriel Moreno-Hagelsieb
Ernesto Pérez-Rueda
author_sort Leonardo Ledesma-Dominguez
collection DOAJ
description Abstract Deep learning models (DLMs) have gained importance in predicting, detecting, translating, and classifying a diversity of inputs. In bioinformatics, DLMs have been used to predict protein structures, transcription factor-binding sites, and promoters. In this work, we propose a hybrid model to identify transcription factors (TFs) among prokaryotic and eukaryotic protein sequences, named Deep Regulation (DeepReg) model. Two architectures were used in the DL model: a convolutional neural network (CNN), and a bidirectional long-short-term memory (BiLSTM). DeepReg reached a precision of 0.99, a recall of 0.97, and an F1-score of 0.98. The quality of our predictions, the bias-variance trade-off approach, and the characterization of new TF predictions were evaluated and compared against those produced by DeepTFactor, as well as against experimental data from three model organisms. Predictions based on our DLM tended to exhibit less variance and bias than those from DeepTFactor, thus increasing reliability and decreasing overfitting.
format Article
id doaj-art-a57e838fde374b9aa17f0e0e5b4b5edf
institution Kabale University
issn 2045-2322
language English
publishDate 2024-04-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-a57e838fde374b9aa17f0e0e5b4b5edf2024-11-17T12:21:24ZengNature PortfolioScientific Reports2045-23222024-04-0114111110.1038/s41598-024-59487-5DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomesLeonardo Ledesma-Dominguez0Erik Carbajal-Degante1Gabriel Moreno-Hagelsieb2Ernesto Pérez-Rueda3Posgrado en Ciencia en Ingeniería de la Computación, Universidad Nacional Autónoma de MéxicoCoordinación de Universidad Abierta, Innovación Educativa y Educación a Distancia (CUAIEED), Universidad Nacional Autónoma de MéxicoDepartment of Biology, Wilfrid Laurier UniversityInstituto de Investigaciones en Matemáticas Aplicadas y en Sistemas, Unidad Académica del Estado de Yucatán, Universidad Nacional Autónoma de MéxicoAbstract Deep learning models (DLMs) have gained importance in predicting, detecting, translating, and classifying a diversity of inputs. In bioinformatics, DLMs have been used to predict protein structures, transcription factor-binding sites, and promoters. In this work, we propose a hybrid model to identify transcription factors (TFs) among prokaryotic and eukaryotic protein sequences, named Deep Regulation (DeepReg) model. Two architectures were used in the DL model: a convolutional neural network (CNN), and a bidirectional long-short-term memory (BiLSTM). DeepReg reached a precision of 0.99, a recall of 0.97, and an F1-score of 0.98. The quality of our predictions, the bias-variance trade-off approach, and the characterization of new TF predictions were evaluated and compared against those produced by DeepTFactor, as well as against experimental data from three model organisms. Predictions based on our DLM tended to exhibit less variance and bias than those from DeepTFactor, thus increasing reliability and decreasing overfitting.https://doi.org/10.1038/s41598-024-59487-5
spellingShingle Leonardo Ledesma-Dominguez
Erik Carbajal-Degante
Gabriel Moreno-Hagelsieb
Ernesto Pérez-Rueda
DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes
Scientific Reports
title DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes
title_full DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes
title_fullStr DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes
title_full_unstemmed DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes
title_short DeepReg: a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes
title_sort deepreg a deep learning hybrid model for predicting transcription factors in eukaryotic and prokaryotic genomes
url https://doi.org/10.1038/s41598-024-59487-5
work_keys_str_mv AT leonardoledesmadominguez deepregadeeplearninghybridmodelforpredictingtranscriptionfactorsineukaryoticandprokaryoticgenomes
AT erikcarbajaldegante deepregadeeplearninghybridmodelforpredictingtranscriptionfactorsineukaryoticandprokaryoticgenomes
AT gabrielmorenohagelsieb deepregadeeplearninghybridmodelforpredictingtranscriptionfactorsineukaryoticandprokaryoticgenomes
AT ernestoperezrueda deepregadeeplearninghybridmodelforpredictingtranscriptionfactorsineukaryoticandprokaryoticgenomes