Machine learning methods accurately predict host specificity of coronaviruses based on spike sequences alone

Kuzmin, K. G.; Adeniyi, Ayotomiwa Ezekiel; DaSouza, Arthur Kevin; Lim, Deuk; Nguyen, H.N.; Molina, Nuria Ramirez; Xiong, Lanqiao; Weber, Irene T.; Harrison, Robert W.

doi:10.1016/j.bbrc.2020.09.010

Cited by 79 publications

(85 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Performing different data analytics tasks on sequences has been done successfully by different researchers previously [20,22]. However, most studies require the sequences to be aligned [10,23,24]. The aligned sequences are used to generate fixed length numerical embeddings, which can then be used for tasks such as classification and clustering [20,25,26].…”

Section: Literature Reviewmentioning

confidence: 99%

“…Due to the availability of large-scale sequence data for the SARS-CoV-2 virus, an accurate and effective clustering method is needed to further analyze this disease, so as to better understand the dynamics and diversity of this virus. To classify different coronavirus hosts, authors in [10] suggest a one-hot encoding-based method that uses spike sequences alone. Their study reveals that they achieved excellent prediction accuracy considering just the spike portion of the genome sequence instead of using the entire sequence.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Their study reveals that they achieved excellent prediction accuracy considering just the spike portion of the genome sequence instead of using the entire sequence. Using this idea and a kernel method, Ali et al in [20] accomplish higher accuracy than in [10], in the classification of different variants of SARS-CoV-2 in humans. Successful analysis of different variants leads to designing efficient strategy regarding the vaccination distribution [41][42][43][44].…”

Section: Literature Reviewmentioning

confidence: 99%

“…The S protein has an essential role in viral infection that makes it a potential target for vaccine development, antibodyblocking therapy, and small molecule inhibitors [9]. In addition, the spike region of the SARS-CoV-2 genome is involved in a disproportionate amount of the genomic variation, for its length [10] (see, e.g., Table 1). Therefore, mutations that affect the antigenicity of the S protein are of certain importance [11].…”

Section: Introductionmentioning

confidence: 99%

“…As a result, these amino acid sequences should be converted into numeric characters in a way that preserves some sequential order information of the amino acids within each sequence. The most prevalent strategy in this area is one-hot encoding due to its simplicity [10]. Since we need to compute pairwise distances (e.g., Euclidean distance), one-hot encoding order preservation would not be operational [20].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Robust Representation and Efficient Feature Selection Allows for Effective Clustering of SARS-CoV-2 Variants

2021

View full text Add to dashboard Cite

The widespread availability of large amounts of genomic data on the SARS-CoV-2 virus, as a result of the COVID-19 pandemic, has created an opportunity for researchers to analyze the disease at a level of detail, unlike any virus before it. On the one hand, this will help biologists, policymakers, and other authorities to make timely and appropriate decisions to control the spread of the coronavirus. On the other hand, such studies will help to more effectively deal with any possible future pandemic. Since the SARS-CoV-2 virus contains different variants, each of them having different mutations, performing any analysis on such data becomes a difficult task, given the size of the data. It is well known that much of the variation in the SARS-CoV-2 genome happens disproportionately in the spike region of the genome sequence—the relatively short region which codes for the spike protein(s). In this paper, we propose a robust feature-vector representation of biological sequences that, when combined with the appropriate feature selection method, allows different downstream clustering approaches to perform well on a variety of different measures. We use such proposed approach with an array of clustering techniques to cluster spike protein sequences in order to study the behavior of different known variants that are increasing at a very high rate throughout the world. We use a k-mers based approach first to generate a fixed-length feature vector representation of the spike sequences. We then show that we can efficiently and effectively cluster the spike sequences based on the different variants with the appropriate feature selection. Using a publicly available set of SARS-CoV-2 spike sequences, we perform clustering of these sequences using both hard and soft clustering methods and show that, with our feature selection methods, we can achieve higher F1 scores for the clusters and also better clustering quality metrics compared to baselines.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Section: Literature Reviewmentioning

confidence: 99%