Infectious diseases caused by novel viruses have become a major public health concern. Rapid identification of virus–host interactions can reveal mechanistic insights into infectious diseases and shed light on potential treatments. Current computational prediction methods for novel viruses are based mainly on protein sequences. However, it is not clear to what extent other important features, such as the symptoms caused by the viruses, could contribute to a predictor. Disease phenotypes (i.e. signs and symptoms) are readily accessible from clinical diagnosis and we hypothesize that they may act as a potential proxy and an additional source of information for the underlying molecular interactions between the pathogens and hosts. We developed DeepViral, a deep learning based method that predicts protein–protein interactions (PPI) between humans and viruses. Motivated by the potential utility of infectious disease phenotypes, we first embedded human proteins and viruses in a shared space using their associated phenotypes and functions, supported by formalized background knowledge from biomedical ontologies. By jointly learning from protein sequences and phenotype features, DeepViral significantly improves over existing sequence-based methods for intra- and inter-species PPI prediction. Code and datasets for reproduction and customization are available at https://github.com/bio-ontology-research-group/DeepViral. Prediction results for 14 virus families are available at https://doi.org/10.5281/zenodo.4429824. Supplementary data are available at Bioinformatics online.
The biases in CoDing Sequence (CDS) prediction tools, which have been based on historic genomic annotations from model organisms, impact our understanding of novel genomes and metagenomes. This hinders the discovery of new genomic information as it results in predictions being biased towards existing knowledge. To date, users have lacked a systematic and replicable approach to identify the strengths and weaknesses of any CDS prediction tool and allow them to choose the right tool for their analysis. We present an evaluation framework (ORForise) based on a comprehensive set of 12 primary and 60 secondary metrics that facilitate the assessment of the performance of CDS prediction tools. This makes it possible to identify which performs better for specific use-cases. We use this to assess 15 ab initio- and model-based tools representing those most widely used (historically and currently) to generate the knowledge in genomic databases. We find that the performance of any tool is dependent on the genome being analysed, and no individual tool ranked as the most accurate across all genomes or metrics analysed. Even the top-ranked tools produced conflicting gene collections, which could not be resolved by aggregation. The ORForise evaluation framework provides users with a replicable, data-led approach to make informed tool choices for novel genome annotations and for refining historical annotations. Code and datasets for reproduction and customisation are available at https://github.com/NickJD/ORForise. Supplementary data are available at Bioinformatics online.
In 2019, a novel coronavirus, SARS-CoV-2/nCoV-19, emerged in Wuhan, China, and has been responsible for the current COVID-19 pandemic. The evolutionary origins of the virus remain elusive and understanding its complex mutational signatures could guide vaccine design and development. As part of the international “CoronaHack” in April 2020, we employed a collection of contemporary methodologies to compare the genomic sequences of coronaviruses isolated from human (SARS-CoV-2; n = 163), bat (bat-CoV; n = 215) and pangolin (pangolin-CoV; n = 7) available in public repositories. We have also noted the pangolin-CoV isolate MP789 to bare stronger resemblance to SARS-CoV-2 than other pangolin-CoV. Following de novo gene annotation prediction, analyses of gene–gene similarity network, codon usage bias and variant discovery were undertaken. Strong host-associated divergences were noted in ORF3a, ORF6, ORF7a, ORF8 and S, and in codon usage bias profiles. Last, we have characterised several high impact variants (in-frame insertion/deletion or stop gain) in bat-CoV and pangolin-CoV populations, some of which are found in the same amino acid position and may be highlighting loci of potential functional relevance.
A deeper understanding of the relationship between the antimicrobial resistance (AMR) gene carriage and phenotype is necessary to develop effective response strategies against this global burden. AMR phenotype is often a result of multi-gene interactions; therefore, we need approaches that go beyond current simple AMR gene identification tools. Machine-learning (ML) methods may meet this challenge and allow the development of rapid computational approaches for AMR phenotype classification. To examine this, we applied multiple ML techniques to 16,950 bacterial genomes across 28 genera, with corresponding MICs for 23 antibiotics with the aim of training models to accurately determine the AMR phenotype from sequenced genomes. This resulted in a >1.5-fold increase in AMR phenotype prediction accuracy over AMR gene identification alone. Furthermore, we revealed 528 unique (often species-specific) genomic routes to antibiotic resistance, including genes not previously linked to the AMR phenotype. Our study demonstrates the utility of ML in predicting AMR phenotypes across diverse clinically relevant organisms and antibiotics. This research proposes a rapid computational method to support laboratory-based identification of the AMR phenotype in pathogens.
Microbes use a range of genetic codes and gene structures, yet these are often ignored during metagenomic analysis. This causes spurious protein predictions, preventing functional assignment which limits our understanding of ecosystems. To resolve this, we developed a lineage-specific gene prediction approach that uses the correct genetic code based on the taxonomic assignment of genetic fragments, removes incomplete protein predictions, and optimises prediction of small proteins. Applied to 9634 metagenomes and 3594 genomes from the human gut, this approach increased the landscape of captured expressed microbial proteins by 78.9%, including previously hidden functional groups. Optimised small protein prediction captured 3,772,658 small protein clusters, which form an improved microbial protein catalogue of the human gut (MiProGut). To enable the ecological study of a protein’s prevalence and association with host parameters, we developed InvestiGUT, a tool which integrates both the protein sequences and sample metadata. Accurate prediction of proteins is critical to providing a functional understanding of microbiomes, enhancing our ability to study interactions between microbes and hosts.
Large regions of prokaryotic genomes are currently without any annotation, in part due to well-established limitations of annotation tools. For example, it is routine for genes using alternative start codons to be misreported or completely omitted. Therefore, we present StORF-Reporter, a tool that takes an annotated genome and returns regions that may contain missing CDS genes from unannotated regions. StORF-Reporter consists of two parts. The first begins with the extraction of unannotated regions from an annotated genome. Next, Stop-ORFs (StORFs) are identified in these unannotated regions. StORFs are open reading frames that are delimited by stop codons and thus can capture those genes most often missing in genome annotations. We show this methodology recovers genes missing from canonical genome annotations. We inspect the results of the genomes of model organisms, the pangenome of Escherichia coli, and a set of 5109 prokaryotic genomes of 247 genera from the Ensembl Bacteria database. StORF-Reporter extended the core, soft-core and accessory gene collections, identified novel gene families and extended families into additional genera. The high levels of sequence conservation observed between genera suggest that many of these StORFs are likely to be functional genes that should now be considered for inclusion in canonical annotations.