Viruses have caused much mortality and morbidity to humans and pose a serious threat to global public health. The virome with the potential of human infection is still far from complete. Novel viruses have been discovered at an unprecedented pace as the rapid development of viral metagenomics. However, there is still a lack of methodology for rapidly identifying novel viruses with the potential of human infection. This study built several machine learning models to discriminate human‐infecting viruses from other viruses based on the frequency of k‐mers in the viral genomic sequences. The k‐nearest neighbor (KNN) model can predict the human‐infecting viruses with an accuracy of over 90%. The performance of this KNN model built on the short contigs (≥1 kb) is comparable to those built on the viral genomes. We used a reported human blood virome to further validate this KNN model with an accuracy of over 80% based on very short raw reads (150 bp). Our work demonstrates a conceptual and generic protocol for the discovery of novel human‐infecting viruses in viral metagenomics studies.
Newly emerging influenza viruses continue to threaten public health. A rapid determination of the host range of newly discovered influenza viruses would assist in early assessment of their risk. Here, we attempted to predict the host of influenza viruses using the Support Vector Machine (SVM) classifier based on the word vector, a new representation and feature extraction method for biological sequences. The results show that the length of the word within the word vector, the sequence type (DNA or protein) and the species from which the sequences were derived for generating the word vector all influence the performance of models in predicting the host of influenza viruses. In nearly all cases, the models built on the surface proteins hemagglutinin (HA) and neuraminidase (NA) (or their genes) produced better results than internal influenza proteins (or their genes). The best performance was achieved when the model was built on the HA gene based on word vectors (words of three-letters long) generated from DNA sequences of the influenza virus. This results in accuracies of 99.7% for avian, 96.9% for human and 90.6% for swine influenza viruses. Compared to the method of sequence homology best-hit searches using the Basic Local Alignment Search Tool (BLAST), the word vector-based models still need further improvements in predicting the host of influenza A viruses.
High-throughput reporter assays have been recently developed to directly and quantitatively assess enhancer activity for thousands of regulatory elements. However, there is still no database to collect these enhancers. We developed RAEdb, the first database to collect enhancers identified by high-throughput reporter assays. RAEdb includes 538 320 enhancers derived from eight studies, most of which were from six human cell lines. An activity score was assigned to each enhancer based on reporter assays. Based on these enhancers, 7658 epromoters (promoters with enhancer activity) were identified and stored in the database. RAEdb provides two ways of searches: the first is to search studies by species and cell line; the other is to search enhancers or epromoters by position, activity score, sequence and gene. RAEdb also provides a genome browser to query, visualize and compare enhancers. All data in RAEdb is freely available for download.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.