Untangling an insect's virome from its endogenous viral elements

detectEVE: Fast, Sensitive and Precise Detection of Endogenous Viral Elements in Genomic Data

Endogenous viral elements (EVEs) are fragments of viral genomic material embedded within the host genome. Retroviruses contribute to the majority of EVEs because of their genomic integration during their life cycle; however, the latter can also arise from non-retroviral RNA or DNA viruses, then collectively known as non-retroviral (nr) EVEs. Detecting nrEVEs poses challenges because of their sequence and genomic structural diversity, contributing to the scarcity of specific tools designed for nrEVEs detection. Here, we introduce detectEVE, a user-friendly and open-source tool designed for the accurate identification of nrEVEs in genomic assemblies. detectEVE deviates from other nrEVE detection pipelines, which usually classify sequences in a more rigid manner as either virus-associated or not. Instead, we implemented a scaling system assigning confidence scores to hits in protein sequence similarity searches, using bit score distributions and search hints related to various viral characteristics, allowing for higher sensitivity and specificity. Our benchmarking shows that detectEVE is computationally efficient and accurate, as well as considerably faster than existing approaches, because of its resource-efficient parallel execution. Our tool can help to fill current gaps in both host-associated fields and virus-related studies. This includes (i) enhancing genome annotations with metadata for EVE loci, (ii) conducting large-scale paleo-virological studies to explore deep viral evolutionary histories, and (iii) aiding in the identification of actively expressed EVEs in transcriptomic data, reducing the risk of misinterpretations between exogenous viruses and EVEs.

ViraLM: empowering virus discovery through the genome foundation model

MOTIVATION

Viruses, with their ubiquitous presence and high diversity, play pivotal roles in ecological systems and public health. Accurate identification of viruses in various ecosystems is essential for comprehending their variety and assessing their ecological influence. Metagenomic sequencing has become a major strategy to survey the viruses in various ecosystems. However, accurate and comprehensive virus detection in metagenomic data remains difficult. Limited reference sequences prevent alignment-based methods from identifying novel viruses. Machine learning-based tools are more promising in novel virus detection but often miss short viral contigs, which are abundant in typical metagenomic data. The inconsistency in virus search results produced by available tools further highlights the urgent need for a more robust tool for virus identification.

RESULTS

In this work, we develop ViraLM for identifying novel viral contigs in metagenomic data. By using the latest genome foundation model as the backbone and training on a rigorously constructed dataset, the model is able to distinguish viruses from other organisms based on the learned genomic characteristics. We thoroughly tested ViraLM on multiple datasets and the experimental results show that ViraLM outperforms available tools in different scenarios. In particular, ViraLM improves the F1-score on short contigs by 22%.

AVAILABILITY AND IMPLEMENTATION

The source code of ViraLM is available via: https://github.com/ChengPENG-wolf/ViraLM.

A tale of caution: How endogenous viral elements affect virus discovery in transcriptomic data

Large-scale metagenomic and -transcriptomic studies have revolutionized our understanding of viral diversity and abundance. In contrast, endogenous viral elements (EVEs), remnants of viral sequences integrated into host genomes, have received limited attention in the context of virus discovery, especially in RNA-Seq data. EVEs resemble their original viruses, a challenge that makes distinguishing between active infections and integrated remnants difficult, affecting virus classification and biases downstream analyses. Here, we systematically assess the effects of EVEs on a prototypical virus discovery pipeline, evaluate their impact on data integrity and classification accuracy, and provide some recommendations for better practices. We examined EVEs and exogenous viral sequences linked to Orthomyxoviridae, a diverse family of negative-sense segmented RNA viruses, in 13 genomic and 538 transcriptomic datasets of Culicinae mosquitoes. Our analysis revealed a substantial number of viral sequences in transcriptomic datasets. However, a significant portion appeared not to be exogenous viruses but transcripts derived from EVEs. Distinguishing between transcribed EVEs and exogenous virus sequences was especially difficult in samples with low viral abundance. For example, three transcribed EVEs showed full-length segments, devoid of frameshift and nonsense mutations, exhibiting sufficient mean read depths that qualify them as exogenous virus hits. Mapping reads on a host genome containing EVEs before assembly somewhat alleviated the EVE burden, but it led to a drastic reduction of viral hits and reduced quality of assemblies, especially in regions of the viral genome relatively similar to EVEs. Our study highlights that our knowledge of the genetic diversity of viruses can be altered by the underestimated presence of EVEs in transcriptomic datasets, leading to false positives and altered or missing sequence information. Thus, recognizing and addressing the influence of EVEs in virus discovery pipelines will be key in enhancing our ability to capture the full spectrum of viral diversity.