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

Randhawa GS, Soltysiak MPM, El Roz H, de Souza CPE, Hill KA, Kari L. · 2020

Read Machine learning using intrinsic genomic signatures for rapid classification of novel pathogens: COVID-19 case study. by Randhawa GS, Soltysiak MPM, El Roz H, de Souza CPE, Hill KA, Kari L. (2020) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc7182198.txt.

Introduction

Randhawa and co-authors test a fast machine learning method for classifying new pathogens. Their study appeared in PLoS ONE. The method reads raw genome sequence only. It needs no gene labels or expert training data. The team used the COVID-19 virus as a test case. Early, correct naming of a new virus guides response plans. Slow methods can delay that work.

The team analyzed more than 5,000 viral genomes. The set held 61.8 million base pairs in total. Their method mixes machine learning with signal processing. A decision tree and a rank test helped check the results. The authors report perfect accuracy on the COVID-19 virus sequences. The method placed the virus among the Sarbecovirus group of betacoronaviruses. It also supported a bat origin. The whole scan took only minutes, they report.

Catalog record

CREATORRandhawa GS, Soltysiak MPM, El Roz H, de Souza CPE, Hill KA, Kari L.
YEAR2020
COLLECTIONPAPERS
FORMATTXT
FILE SIZE66 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC7182198

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Machine learning using intrinsic genomic signatures for rapid classification of novel pathogens: COVID-19 case study By Gurjit S Randhawa, Maximillian P M Soltysiak, Hadi El Roz, Camila P E de Souza, Kathleen A Hill, Lila Kari, Oliver Schildgen Published in: PLoS ONE DOI: 10.1371/journal.pone.0232391 ABSTRACT Abstract 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. INTRODUCTION Coronaviruses are single-stranded positive-sense RNA viruses that are known to contain some of the largest viral genomes, up to around 32 kbp in length [1–5]. After increases in the number of coronavirus genome sequences available following efforts to investigate the diversity in the wild, the family Coronaviridae now contains four genera (International Committee on Taxonomy of Viruses, [6]). While those species that belong to the genera Alphacoronavirus and Betacoronavirus can infect mammalian hosts, those in Gammacoronavirus and the recently defined Deltacoronavirus mainly infect avian species [4, 7–9]. Phylogenetic studies have revealed a complex evolutionary history, with coronaviruses thought to have ancient origins and recent crossover events that can lead to cross-species infection [8, 10–12]. Some of the largest sources of diversity for coronaviruses belong to the strains that infect bats and birds, providing a reservoir in wild animals for recombination and mutation that may enable cross-species transmission into other mammals and humans [4, 7, 8, 10, 13]. Like other RNA viruses, coronavirus genomes are known to have genomic plasticity, and this can be attributed to several major factors. RNA-dependent RNA polymerases (RdRp) have high mutation rates, reaching from 1 in 1000 to 1 in 10000 nucleotides during replication [7, 14, 15]. Coronaviruses are also known to use a template switching mechanism which can contribute to high rates of homologous RNA recombination between their viral genomes [9, 16–20]. Furthermore, the large size of coronavirus genomes is thought to be able to accommodate mutations to genes [7]. These factors help contribute to the plasticity and diversity of coronavirus genomes today. The highly pathogenic human coronaviruses, Severe Acute Respiratory Syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV) belong to lineage B (sub-genus Sarbecovirus) and lineage C (sub-genus Merbecovirus) of Betacoronavirus, respectively [9, 21–23]. Both result from zoonotic transmission to humans and lead to symptoms of viral pneumonia, including fever, breathing difficulties, and more [24, 25]. Recently, an unidentified pneumonia disease with similar symptoms caused an outbreak in Wuhan and is thought to have started from a local fresh seafood market [26–30]. This was later attributed to a novel coronavirus (the COVID-19 virus), and represents the third major zoonotic human coronavirus of this century [31]: On February 28, 2020, the World Health Organization set the COVID-19 risk assessment for regional and global levels to “Very High” [32]. From analyses employing whole genome to viral protein-based comparisons, the COVID-19 virus is thought to belong to lineage B (Sarbecovirus) of Betacoronavirus. From phylogenetic analysis of the RdRp protein, spike proteins, and full genomes of the COVID-19 virus and other coronaviruses, it was found that the COVID-19 virus is most closely related to two bat SARS-like coronaviruses, bat-SL-CoVZXC21 and bat-SL-CoVZC45, found in Chinese horseshoe bats Rhinolophus sinicus [12, 33–37]. Along with the phylogenetic data, the genome organization of the COVID-19 virus was found to be typical of lineage B (Sarbecovirus) Betacoronaviruses [33]. From phylogenetic analysis of full genome alignment and similarity plots, it was found that the COVID-19 virus has the highest similarity to the bat coronavirus RaTG13 [38]. Close associations to bat coronavirus RaTG13 and two bat SARS-like CoVs (ZC45 and ZXC21) are also supported in alignment-based phylogenetic analyses [38]. Within the COVID-19 virus sequences, over 99% sequence similarity and a lack of diversity within these strains suggest a common lineage and source, with support for recent emergence of the human strain [12, 31]. There is ongoing debate whether the COVID-19 virus arose following recombination with previously identified bat and unknown coronaviruses [39] or arose independently as a new lineage to infect humans [38]. In combination with the identification that the angiotensin converting enzyme 2 (ACE2) protein is a receptor for COVID-19 virus, as it is for SARS and other Sarbecovirus strains, the hypothesis that the COVID-19 virus originated from bats is deemed very likely [12, 33, 35, 38, 40–44]. All analyses performed thus far have been alignment-based and rely on the annotations of the viral genes. Though alignment-based methods have been successful in finding sequence similarities, their application can be challenging in many cases [45, 46]. It is realistically impossible to analyze thousands of complete genomes using alignment-based methods due to the heavy computation time. Moreover, the alignment demands the sequences to be continuously homologous which is not always the case. Alignment-free methods [47–51] have been proposed in the past as an alternative to address the limitations of the alignment-based methods. Comparative genomics beyond alignment-based approaches have benefited from the computational power of machine learning. Machine learning-based alignment-free methods have also been used successfully for a variety of problems including virus classification [49–51]. An alignment-free approach [49] was proposed for subtype classification of HIV-1 genomes and achieved ∼97% classification accuracy. MLDSP [50], with the use of a broad range of 1D numerical representations of DNA sequences, has also achieved very high levels of classification accuracy with viruses. Even rapidly evolving, plastic genomes of viruses such as Influenza and Dengue are classified down to the level of strain and subtype, respectively with 100% classification accuracy. MLDSP-GUI [51] provides an option to use 2D Chaos Game Representation (CGR) [52] as numerical representation of DNA sequences. CGR’s have a longstanding use in species classification with identification of biases in sequence composition [48, 51, 52]. MLDSP-GUI has shown 100% classification accuracy for Flavivirus genus to species classification using 2D CGR as numerical representation [51]. MLDSP and MLDSP-GUI have demonstrated the ability to identify the genomic signatures (a species-specific pattern known to be pervasive throughout the genome) with species level accuracy that can be used for sequence

Read the complete text