Analysis of DNA Sequence Classification Using CNN and Hybrid Models

Abstract

In a general computational context for biomedical data analysis, DNA sequence classification is a crucial challenge. Several machine learning techniques have used to complete this task in recent years successfully. Identification and classification of viruses are essential to avoid an outbreak like COVID-19. Regardless, the feature selection process remains the most challenging aspect of the issue. The most commonly used representations worsen the case of high dimensionality, and sequences lack explicit features. It also helps in detecting the effect of viruses and drug design. In recent days, deep learning (DL) models can automatically extract the features from the input. In this work, we employed CNN, CNN-LSTM, and CNN-Bidirectional LSTM architectures using Label and K -mer encoding for DNA sequence classification. The models are evaluated on different classification metrics. From the experimental results, the CNN and CNN-Bidirectional LSTM with K -mer encoding offers high accuracy with 93.16% and 93.13%, respectively, on testing data.

A Deep Learning Approach for Viral DNA Sequence Classification using Genetic Algorithm

“…The error between the actual output and the goal label, on which the weights are trained and updated, is calculated using the loss function of the GA algorithm. A variety of hyperparameters, such as filter size, layer count, and embedding dimension, were used to evaluate the CNN, CNN-LSTM, and CNN-bidirectional LSTM models but the same architecture is used and the same hyperparameters as [18] in testing and evaluation to correctly compare the results. The embedding layer has 8 dimensions, which is the initial layer.…”

“…The same LSTM and LSTM/CNN hybrid models are used in [18] to correctly compare the results and improve upon the currently existing model after adding the GA layer and using ADASYN for oversampling as well as increasing the dataset variability. Increasing the number of class labels in the dataset and the number of input sequences also contributed to the overall better performance of the models.…”

“…Increasing the number of class labels in the dataset and the number of input sequences also contributed to the overall better performance of the models. The accuracy increased as compared to [18] by the introduction of the two new class labels for the Zika and the Ebola virus as well as the additional data collected for the rest of the class labels. Label encoding achieved the best accuracy in the CNN classification layer in both training and testing thus it would achieve the best results in the remaining layers.…”

“…In [18], the author proposed three models for the classification of different viral DNA sequences using raw DNA sequence data. The three classification models were CNN, long short-term memory (LSTM), and convolutional neural network bidirectional long short-term memory (CNN-Bidirectional LSTM).…”

Analysis and Allocation of Cancer-Related Genes Using Vague DNA Sequence Data

“… Yao, Jin, & Lee (2018 ) improve the statistical analysis for genetic data. Gunasekaran et al (2021 ) analyze DNA data using hybrid models. Halla-aho and Lähdesmäki (2021 ) use statistical analysis for DNA cancer data.…”