An investigation into inter- and intragenomic variations of graphic genomic signatures

On leveraging self-supervised learning for accurate HCV genotyping

Hepatitis C virus (HCV) is a major global health concern, affecting millions of individuals worldwide. While existing literature predominantly focuses on disease classification using clinical data, there exists a critical research gap concerning HCV genotyping based on genomic sequences. Accurate HCV genotyping is essential for patient management and treatment decisions. While the neural models excel at capturing complex patterns, they still face challenges, such as data scarcity, that exist a lot in computational genomics. To overcome this challenges, this paper introduces an advanced deep learning approach for HCV genotyping based on the graphical representation of nucleotide sequences that outperforms classical approaches. Notably, it is effective for both partial and complete HCV genomes and addresses challenges associated with imbalanced datasets. In this work, ten HCV genotypes: 1a, 1b, 2a, 2b, 2c, 3a, 3b, 4, 5, and 6 were used in the analysis. This study utilizes Chaos Game Representation for 2D mapping of genomic sequences, employing self-supervised learning using convolutional autoencoder for deep feature extraction, resulting in an outstanding performance for HCV genotyping compared to various machine learning and deep learning models. This baseline provides a benchmark against which the performance of the proposed approach and other models can be evaluated. The experimental results showcase a remarkable classification accuracy of over 99%, outperforming traditional deep learning models. This performance demonstrates the capability of the proposed model to accurately identify HCV genotypes in both partial and complete sequences and in dealing with data scarcity for certain genotypes. The results of the proposed model are compared to NCBI genotyping tool.

Automated classification of giant virus genomes using a random forest model built on trademark protein families

Viruses of the phylum Nucleocytoviricota, often referred to as "giant viruses," are prevalent in various environments around the globe and play significant roles in shaping eukaryotic diversity and activities in global ecosystems. Given the extensive phylogenetic diversity within this viral group and the highly complex composition of their genomes, taxonomic classification of giant viruses, particularly incomplete metagenome-assembled genomes (MAGs) can present a considerable challenge. Here we developed TIGTOG (Taxonomic Information of Giant viruses using Trademark Orthologous Groups), a machine learning-based approach to predict the taxonomic classification of novel giant virus MAGs based on profiles of protein family content. We applied a random forest algorithm to a training set of 1531 quality-checked, phylogenetically diverse Nucleocytoviricota genomes using pre-selected sets of giant virus orthologous groups (GVOGs). The classification models were predictive of viral taxonomic assignments with a cross-validation accuracy of 99.6% at the order level and 97.3% at the family level. We found that no individual GVOGs or genome features significantly influenced the algorithm's performance or the models' predictions, indicating that classification predictions were based on a comprehensive genomic signature, which reduced the necessity of a fixed set of marker genes for taxonomic assigning purposes. Our classification models were validated with an independent test set of 823 giant virus genomes with varied genomic completeness and taxonomy and demonstrated an accuracy of 98.6% and 95.9% at the order and family level, respectively. Our results indicate that protein family profiles can be used to accurately classify large DNA viruses at different taxonomic levels and provide a fast and accurate method for the classification of giant viruses. This approach could easily be adapted to other viral groups.

Genomic Signature in Evolutionary Biology: A Review

Organisms are unique physical entities in which information is stored and continuously processed. The digital nature of DNA sequences enables the construction of a dynamic information reservoir. However, the distinction between the hardware and software components in the information flow is crucial to identify the mechanisms generating specific genomic signatures. In this work, we perform a bibliometric analysis to identify the different purposes of looking for particular patterns in DNA sequences associated with a given phenotype. This study has enabled us to make a conceptual breakdown of the genomic signature and differentiate the leading applications. On the one hand, it refers to gene expression profiling associated with a biological function, which may be shared across taxa. This signature is the focus of study in precision medicine. On the other hand, it also refers to characteristic patterns in species-specific DNA sequences. This interpretation plays a key role in comparative genomics, identifying evolutionary relationships. Looking at the relevant studies in our bibliographic database, we highlight the main factors causing heterogeneities in genome composition and how they can be quantified. All these findings lead us to reformulate some questions relevant to evolutionary biology.

Chaos game representation and its applications in bioinformatics

Chaos game representation (CGR), a milestone in graphical bioinformatics, has become a powerful tool regarding alignment-free sequence comparison and feature encoding for machine learning. The algorithm maps a sequence to 2-dimensional space, while an extension of the CGR, the so-called frequency matrix representation (FCGR), transforms sequences of different lengths into equal-sized images or matrices. The CGR is a generalized Markov chain and includes various properties, which allow a unique representation of a sequence. Therefore, it has a broad spectrum of applications in bioinformatics, such as sequence comparison and phylogenetic analysis and as an encoding of sequences for machine learning. This review introduces the construction of CGRs and FCGRs, their applications on DNA and proteins, and gives an overview of recent applications and progress in bioinformatics.

Alignment free sequence comparison methods and reservoir host prediction

Motivation

The emergence and subsequent pandemic of the SARS-CoV-2 virus raised urgent questions about its origin and, particularly, its reservoir host. These types of questions are long-standing problems in the management of emerging infectious diseases and are linked to virus discovery programs and the prediction of viruses that are likely to become zoonotic. Conventional means to identify reservoir hosts have relied on surveillance, experimental studies and phylogenetics. More recently, machine learning approaches have been applied to generate tools to swiftly predict reservoir hosts from sequence data.

Results

Here, we extend a recent work that combined sequence alignment and a mixture of alignment-free approaches using a gradient boosting machines (GBMs) machine learning model, which integrates genomic traits (GT) and phylogenetic neighbourhood (PN) signatures to predict reservoir hosts. We add a more uniform approach by applying Machine Learning with Digital Signal Processing (MLDSP)-based structural patterns (M-SP). The extended model was applied to an existing virus/reservoir host dataset and to the SARS-CoV-2 and related viruses and generated an improvement in prediction accuracy.

Availability and implementation

The source code used in this work is freely available at https://github.com/bill1167/hostgbms.

Supplementary information

Supplementary data are available at Bioinformatics online.

Similarity Studies of Corona Viruses through Chaos Game Representation

The novel coronavirus (SARS-COV-2) is generally referred to as Covid-19 virus has spread to 213 countries with nearly 7 million confirmed cases and nearly 400,000 deaths. Such major outbreaks demand classification and origin of the virus genomic sequence, for planning, containment, and treatment. Motivated by the above need, we report two alignment-free methods combing with CGR to perform clustering analysis and create a phylogenetic tree based on it. To each DNA sequence we associate a matrix then define distance between two DNA sequences to be the distance between their associated matrix. These methods are being used for phylogenetic analysis of coronavirus sequences. Our approach provides a powerful tool for analyzing and annotating genomes and their phylogenetic relationships. We also compare our tool to ClustalX algorithm which is one of the most popular alignment methods. Our alignment-free methods are shown to be capable of finding closest genetic relatives of coronaviruses.

Machine learning using intrinsic genomic signatures for rapid classification of novel pathogens: COVID-19 case study

The 2019 novel coronavirus (renamed SARS-CoV-2, and generally referred to as the COVID-19 virus) has spread to 184 countries with over 1.5 million confirmed cases. Such major viral outbreaks demand early elucidation of taxonomic classification and origin of the virus genomic sequence, for strategic planning, containment, and treatment. This paper identifies an intrinsic COVID-19 virus genomic signature and uses it together with a machine learning-based alignment-free approach for an ultra-fast, scalable, and highly accurate classification of whole COVID-19 virus genomes. The proposed method combines supervised machine learning with digital signal processing (MLDSP) for genome analyses, augmented by a decision tree approach to the machine learning component, and a Spearman's rank correlation coefficient analysis for result validation. These tools are used to analyze a large dataset of over 5000 unique viral genomic sequences, totalling 61.8 million bp, including the 29 COVID-19 virus sequences available on January 27, 2020. Our results support a hypothesis of a bat origin and classify the COVID-19 virus as Sarbecovirus, within Betacoronavirus. Our method achieves 100% accurate classification of the COVID-19 virus sequences, and discovers the most relevant relationships among over 5000 viral genomes within a few minutes, ab initio, using raw DNA sequence data alone, and without any specialized biological knowledge, training, gene or genome annotations. This suggests that, for novel viral and pathogen genome sequences, this alignment-free whole-genome machine-learning approach can provide a reliable real-time option for taxonomic classification.

Alignment-free genomic sequence comparison using FCGR and signal processing

Background

Alignment-free methods of genomic comparison offer the possibility of scaling to large data sets of nucleotide sequences comprised of several thousand or more base pairs. Such methods can be used for purposes of deducing “nearby” species in a reference data set, or for constructing phylogenetic trees.

Results

We describe one such method that gives quite strong results. We use the Frequency Chaos Game Representation (FCGR) to create images from such sequences, We then reduce dimension, first using a Fourier trig transform, followed by a Singular Values Decomposition (SVD). This gives vectors of modest length. These in turn are used for fast sequence lookup, construction of phylogenetic trees, and classification of virus genomic data. We illustrate the accuracy and scalability of this approach on several benchmark test sets.

Conclusions

The tandem of FCGR and dimension reductions using Fourier-type transforms and SVD provides a powerful approach for alignment-free genomic comparison. Results compare favorably and often surpass best results reported in prior literature. Good scalability is also observed.

MLDSP-GUI: an alignment-free standalone tool with an interactive graphical user interface for DNA sequence comparison and analysis

Summary

Machine Learning with Digital Signal Processing and Graphical User Interface (MLDSP-GUI) is an open-source, alignment-free, ultrafast, computationally lightweight, and standalone software tool with an interactive GUI for comparison and analysis of DNA sequences. MLDSP-GUI is a general-purpose tool that can be used for a variety of applications such as taxonomic classification, disease classification, virus subtype classification, evolutionary analyses, among others.

Availability and implementation

MLDSP-GUI is open-source, cross-platform compatible, and is available under the terms of the Creative Commons Attribution 4.0 International license (http://creativecommons.org/licenses/by/4.0/). The executable and dataset files are available at https://sourceforge.net/projects/mldsp-gui/.

Supplementary information

Supplementary data are available at Bioinformatics online.

Superposition method for modelling boundaries between media in viscoelastic finite difference time domain simulations

Finite-difference time domain (FDTD) techniques are widely used to model the propagation of viscoelastic waves through complex and heterogeneous structures. However, in the specific case of media mixing liquid and solid, attempts to model continuous media onto a Cartesian grid produces errors when the liquid-solid interface between different media do not align precisely with the Cartesian grid. The increase in spatial resolution required to eliminate this grid staircasing effect can be computationally prohibitive. Here, a modification to the Virieux staggered-grid FDTD scheme called the superposition method is presented. This method is intended to reduce this staircasing effect while keeping a manageable computational time. The method was validated by comparing low-spatial-resolution simulations against simulations with sufficiently high resolution to provide reasonably accurate results at any incident angle. The comparison of the root-mean-square of the stress amplitude maps showed that the amplitude of artifactual waves could be reduced by several orders of magnitude when compared to the Virieux staggered-grid FDTD method and that the superposition method helped to significantly reduce the staircasing effect in FDTD simulations.

An open-source k-mer based machine learning tool for fast and accurate subtyping of HIV-1 genomes

For many disease-causing virus species, global diversity is clustered into a taxonomy of subtypes with clinical significance. In particular, the classification of infections among the subtypes of human immunodeficiency virus type 1 (HIV-1) is a routine component of clinical management, and there are now many classification algorithms available for this purpose. Although several of these algorithms are similar in accuracy and speed, the majority are proprietary and require laboratories to transmit HIV-1 sequence data over the network to remote servers. This potentially exposes sensitive patient data to unauthorized access, and makes it impossible to determine how classifications are made and to maintain the data provenance of clinical bioinformatic workflows. We propose an open-source supervised and alignment-free subtyping method (Kameris) that operates on k-mer frequencies in HIV-1 sequences. We performed a detailed study of the accuracy and performance of subtype classification in comparison to four state-of-the-art programs. Based on our testing data set of manually curated real-world HIV-1 sequences (n = 2, 784), Kameris obtained an overall accuracy of 97%, which matches or exceeds all other tested software, with a processing rate of over 1,500 sequences per second. Furthermore, our fully standalone general-purpose software provides key advantages in terms of data security and privacy, transparency and reproducibility. Finally, we show that our method is readily adaptable to subtype classification of other viruses including dengue, influenza A, and hepatitis B and C virus.

MoDMaps3D: an interactive webtool for the quantification and 3D visualization of interrelationships in a dataset of DNA sequences

Summary

MoDMaps3D (Molecular Distance Maps 3D) is an alignment-free, fast, computationally lightweight webtool for computing and visualizing the interrelationships within any dataset of DNA sequences, based on pairwise comparisons between their oligomer compositions. MoDMaps3D is a general-purpose interactive webtool that is free of any requirements on sequence composition, position of the sequences in their respective genomes, presence or absence of similarity or homology, sequence length, or even sequence origin (biological or computer-generated).

Availability and implementation

MoDMaps3D is open source, cross-platform compatible, and is available under the MIT license at http://moleculardistancemaps.github.io/MoDMaps3D/. The source code is available at https://github.com/moleculardistancemaps/MoDMaps3D/.

Contact

lila@uwaterloo.ca.

Supplementary information

Supplementary data are available at Bioinformatics online.

Additive methods for genomic signatures

Background

Studies exploring the potential of Chaos Game Representations (CGR) of genomic sequences to act as “genomic signatures” (to be species- and genome-specific) showed that CGR patterns of nuclear and organellar DNA sequences of the same organism can be very different. While the hypothesis that CGRs of mitochondrial DNA sequences can act as genomic signatures was validated for a snapshot of all sequenced mitochondrial genomes available in the NCBI GenBank sequence database, to our knowledge no such extensive analysis of CGRs of nuclear DNA sequences exists to date.

Results

We analyzed an extensive dataset, totalling 1.45 gigabase pairs, of nuclear/nucleoid genomic sequences (nDNA) from 42 different organisms, spanning all major kingdoms of life. Our computational experiments indicate that CGR signatures of nDNA of two different origins cannot always be differentiated, especially if they originate from closely-related species such as H. sapiens and P. troglodytes or E. coli and E. fergusonii. To address this issue, we propose the general concept of additive DNA signature of a set (collection) of DNA sequences. One particular instance, the composite DNA signature, combines information from nDNA fragments and organellar (mitochondrial, chloroplast, or plasmid) genomes. We demonstrate that, in this dataset, composite DNA signatures originating from two different organisms can be differentiated in all cases, including those where the use of CGR signatures of nDNA failed or was inconclusive. Another instance, the assembled DNA signature, combines information from many short DNA subfragments (e.g., 100 basepairs) of a given DNA fragment, to produce its signature. We show that an assembled DNA signature has the same distinguishing power as a conventionally computed CGR signature, while using shorter contiguous sequences and potentially less sequence information.

Conclusions

Our results suggest that, while CGR signatures of nDNA cannot always play the role of genomic signatures, composite and assembled DNA signatures (separately or in combination) could potentially be used instead. Such additive signatures could be used, e.g., with raw unassembled next-generation sequencing (NGS) read data, when high-quality sequencing data is not available, or to complement information obtained by other methods of species identification or classification.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-016-1157-8) contains supplementary material, which is available to authorized users.