Dataset Information

An improved deep learning method for predicting DNA-binding proteins based on contextual features in amino acid sequences.

ABSTRACT: As the number of known proteins has expanded, how to accurately identify DNA binding proteins has become a significant biological challenge. At present, various computational methods have been proposed to recognize DNA-binding proteins from only amino acid sequences, such as SVM, DNABP and CNN-RNN. However, these methods do not consider the context in amino acid sequences, which makes it difficult for them to adequately capture sequence features. In this study, a new method that coordinates a bidirectional long-term memory recurrent neural network and a convolutional neural network, called CNN-BiLSTM, is proposed to identify DNA binding proteins. The CNN-BiLSTM model can explore the potential contextual relationships of amino acid sequences and obtain more features than can traditional models. The experimental results show that the CNN-BiLSTM achieves a validation set prediction accuracy of 96.5%-7.8% higher than that of SVM, 9.6% higher than that of DNABP and 3.7% higher than that of CNN-RNN. After testing on 20,000 independent samples provided by UniProt that were not involved in model training, the accuracy of CNN-BiLSTM reached 94.5%-12% higher than that of SVM, 4.9% higher than that of DNABP and 4% higher than that of CNN-RNN. We visualized and compared the model training process of CNN-BiLSTM with that of CNN-RNN and found that the former is capable of better generalization from the training dataset, showing that CNN-BiLSTM has a wider range of adaptations to protein sequences. On the test set, CNN-BiLSTM has better credibility because its predicted scores are closer to the sample labels than are those of CNN-RNN. Therefore, the proposed CNN-BiLSTM is a more powerful method for identifying DNA-binding proteins.

SUBMITTER: Hu S

PROVIDER: S-EPMC6855455 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

An improved deep learning method for predicting DNA-binding proteins based on contextual features in amino acid sequences.

Hu Siquan S Ma Ruixiong R Wang Haiou H

PloS one 20191114 11

As the number of known proteins has expanded, how to accurately identify DNA binding proteins has become a significant biological challenge. At present, various computational methods have been proposed to recognize DNA-binding proteins from only amino acid sequences, such as SVM, DNABP and CNN-RNN. However, these methods do not consider the context in amino acid sequences, which makes it difficult for them to adequately capture sequence features. In this study, a new method that coordinates a bi ...[more]

PMID: 31725778

Similar Datasets

Project description:Fusarium head blight (FHB) incited by Fusarium graminearum Schwabe is a devastating disease of barley and other cereal crops worldwide. Fusarium head blight is associated with trichothecene mycotoxins such as deoxynivalenol (DON), where contaminated grains are unfit for malting or animal feed industries. While genetically resistant cultivars offer the best economic and environmentally responsible means to mitigate disease, parent lines with adequate resistance are limited in barley. Resistancebreeding based upon quantitative genetic gains has been slow to date, due to intensive labour requirements of disease nurseries. The development of high throughput genome-wide molecular markers, allow application in genomic prediction models. A diverse genomic panel consisting of 400 two-row spring barley lines was assembled to focus on Canadian barley breeding programs. The panel was evaluated for FHB and DON content in three environments and over two years. Moreover, it was genotyped using an Illumina Infinium HTS iSelect custom beadchip array of single nucleotide polymorphic molecular markers (50K SNP), where over 23K molecular markers were polymorphic. Genomic prediction has been successfully demonstrated for reducing FHB and DON content in cereals using various statistically-based models of different underlying assumptions. Herein, we have studied an alternative method basedon machine learning and compare it with a statistical approach. Two encoding techniques were utilized (categorical or Hardy-Weinberg frequencies), followed by selecting essential genomic markers for phenotype prediction. Subsequently, we applied a transformer-based deep learning algorithm to predict FHB and DON. Apart from the transformer method, we also implemented a Residual Fully Connected Neural Network (RFCNN). Pearson correlation coefficients were calculated to compare true vs. predicted outputs. Under most model scenarios, the use of all markers vs. selected markers marginally improved prediction performance except for RFCNN method for FHB (27.6%). Hardy-Weinberg encoding generally improved correlation for FHB (6.9%) and DON (9.6%) for transformer. This study suggests the potential of the transformer based method for genomic prediction of complex traits such as FHB or DON, having performed better or equally compared with existing machine learning and statistical method. To genomic prediction in barley for Fusarium head blight and deoxynivalenol content using a custom Illumina Infinium array (BarleySNP50-JHI) (www.illumina.com). Sample types included leaves from 400 barley genotypes mostly of Canadian origin. This series includes 400 genotypes assayed on an Illumina infinium HTS platform 50K BeadChip.

Project description:MOTIVATION:Alternative splicing generates multiple isoforms from a single gene, greatly increasing the functional diversity of a genome. Although gene functions have been well studied, little is known about the specific functions of isoforms, making accurate prediction of isoform functions highly desirable. However, the existing approaches to predicting isoform functions are far from satisfactory due to at least two reasons: (i) unlike genes, isoform-level functional annotations are scarce. (ii) The information of isoform functions is concealed in various types of data including isoform sequences, co-expression relationship among isoforms, etc. RESULTS:In this study, we present a novel approach, DIFFUSE (Deep learning-based prediction of IsoForm FUnctions from Sequences and Expression), to predict isoform functions. To integrate various types of data, our approach adopts a hybrid framework by first using a deep neural network (DNN) to predict the functions of isoforms from their genomic sequences and then refining the prediction using a conditional random field (CRF) based on co-expression relationship. To overcome the lack of isoform-level ground truth labels, we further propose an iterative semi-supervised learning algorithm to train both the DNN and CRF together. Our extensive computational experiments demonstrate that DIFFUSE could effectively predict the functions of isoforms and genes. It achieves an average area under the receiver operating characteristics curve of 0.840 and area under the precision-recall curve of 0.581 over 4184 GO functional categories, which are significantly higher than the state-of-the-art methods. We further validate the prediction results by analyzing the correlation between functional similarity, sequence similarity, expression similarity and structural similarity, as well as the consistency between the predicted functions and some well-studied functional features of isoform sequences. AVAILABILITY AND IMPLEMENTATION:https://github.com/haochenucr/DIFFUSE. SUPPLEMENTARY INFORMATION:Supplementary data are available at Bioinformatics online.

Project description:BackgroundThe expansion of raw protein sequence databases in the post genomic era and availability of fresh annotated sequences for major localizations particularly motivated us to introduce a new improved version of our previously forged eukaryotic subcellular localizations prediction method namely "ESLpred". Since, subcellular localization of a protein offers essential clues about its functioning, hence, availability of localization predictor would definitely aid and expedite the protein deciphering studies. However, robustness of a predictor is highly dependent on the superiority of dataset and extracted protein attributes; hence, it becomes imperative to improve the performance of presently available method using latest dataset and crucial input features.ResultsHere, we describe augmentation in the prediction performance obtained for our most popular ESLpred method using new crucial features as an input to Support Vector Machine (SVM). In addition, recently available, highly non-redundant dataset encompassing three kingdoms specific protein sequence sets; 1198 fungi sequences, 2597 from animal and 491 plant sequences were also included in the present study. First, using the evolutionary information in the form of profile composition along with whole and N-terminal sequence composition as an input feature vector of 440 dimensions, overall accuracies of 72.7, 75.8 and 74.5% were achieved respectively after five-fold cross-validation. Further, enhancement in performance was observed when similarity search based results were coupled with whole and N-terminal sequence composition along with profile composition by yielding overall accuracies of 75.9, 80.8, 76.6% respectively; best accuracies reported till date on the same datasets.ConclusionThese results provide confidence about the reliability and accurate prediction of SVM modules generated in the present study using sequence and profile compositions along with similarity search based results. The presently developed modules are implemented as web server "ESLpred2" available at http://www.imtech.res.in/raghava/eslpred2/.

Dataset Information

An improved deep learning method for predicting DNA-binding proteins based on contextual features in amino acid sequences.

Publications

An improved deep learning method for predicting DNA-binding proteins based on contextual features in amino acid sequences.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets