Rfam is a database of RNA families where each of the 3444 families is represented by a multiple sequence alignment of known RNA sequences and a covariance model that can be used to search for additional members of the family. Recent developments have involved expert collaborations to improve the quality and coverage of Rfam data, focusing on microRNAs, viral and bacterial RNAs. We have completed the first phase of synchronising microRNA families in Rfam and miRBase, creating 356 new Rfam families and updating 40. We established a procedure for comprehensive annotation of viral RNA families starting with Flavivirus and Coronaviridae RNAs. We have also increased the coverage of bacterial and metagenome-based RNA families from the ZWD database. These developments have enabled a significant growth of the database, with the addition of 759 new families in Rfam 14. To facilitate further community contribution to Rfam, expert users are now able to build and submit new families using the newly developed Rfam Cloud family curation system. New Rfam website features include a new sequence similarity search powered by RNAcentral, as well as search and visualisation of families with pseudoknots. Rfam is freely available at https://rfam.org.

The Rfam database, a widely used repository of non-coding RNA families, has undergone significant updates in release 15.0. This paper introduces major improvements, including the expansion of Rfamseq to 26 106 genomes, a 76% increase, incorporating the latest UniProt reference proteomes and additional viral genomes. Sixty-five RNA families were enhanced using experimentally determined 3D structures, improving the accuracy of consensus secondary structures and annotations. R-scape covariation analysis was used to refine structural predictions in 26 families. Gene Ontology (GO) and Sequence Ontology annotations were comprehensively updated, increasing GO term coverage to 75% of families. The release adds 14 new Hepatitis C Virus RNA families and completes microRNA family synchronization with miRBase, resulting in 1603 microRNA families. New data types, including FULL alignments, have been implemented. Integration with APICURON for improved curator attribution and multiple website enhancements further improve user experience. These updates significantly expand Rfam’s coverage and improve annotation quality, reinforcing its critical role in RNA research, genome annotation and the development of machine learning models. Rfam is freely available at https://rfam.org.

In recent years, massively parallel complementary DNA sequencing (RNA sequencing [RNA-Seq]) has emerged as a fast, cost-effective, and robust technology to study entire transcriptomes in various manners. In particular, for non-model organisms and in the absence of an appropriate reference genome, RNA-Seq is used to reconstruct the transcriptome de novo. Although the de novo transcriptome assembly of non-model organisms has been on the rise recently and new tools are frequently developing, there is still a knowledge gap about which assembly software should be used to build a comprehensive de novo assembly. Here, we present a large-scale comparative study in which 10 de novo assembly tools are applied to 9 RNA-Seq data sets spanning different kingdoms of life. Overall, we built >200 single assemblies and evaluated their performance on a combination of 20 biological-based and reference-free metrics. Our study is accompanied by a comprehensive and extensible Electronic Supplement that summarizes all data sets, assembly execution instructions, and evaluation results. Trinity, SPAdes, and Trans-ABySS, followed by Bridger and SOAPdenovo-Trans, generally outperformed the other tools compared. Moreover, we observed species-specific differences in the performance of each assembler. No tool delivered the best results for all data sets. We recommend a careful choice and normalization of evaluation metrics to select the best assembling results as a critical step in the reconstruction of a comprehensive de novo transcriptome assembly.

Sequence analyses of RNA virus genomes remain challenging owing to the exceptional genetic plasticity of these viruses. Because of high mutation and recombination rates, genome replication by viral RNA-dependent RNA polymerases leads to populations of closely related viruses, so-called “quasispecies.” Standard (short-read) sequencing technologies are ill-suited to reconstruct large numbers of full-length haplotypes of (1) RNA virus genomes and (2) subgenome-length (sg) RNAs composed of noncontiguous genome regions. Here, we used a full-length, direct RNA sequencing (DRS) approach based on nanopores to characterize viral RNAs produced in cells infected with a human coronavirus. By using DRS, we were able to map the longest (∼26-kb) contiguous read to the viral reference genome. By combining Illumina and Oxford Nanopore sequencing, we reconstructed a highly accurate consensus sequence of the human coronavirus (HCoV)-229E genome (27.3 kb). Furthermore, by using long reads that did not require an assembly step, we were able to identify, in infected cells, diverse and novel HCoV-229E sg RNAs that remain to be characterized. Also, the DRS approach, which circumvents reverse transcription and amplification of RNA, allowed us to detect methylation sites in viral RNAs. Our work paves the way for haplotype-based analyses of viral quasispecies by showing the feasibility of intra-sample haplotype separation. Even though several technical challenges remain to be addressed to exploit the potential of the nanopore technology fully, our work illustrates that DRS may significantly advance genomic studies of complex virus populations, including predictions on long-range interactions in individual full-length viral RNA haplotypes.

Background: The highly diverseCand. Patescibacteria are predicted to have minimal biosynthetic and metabolic pathways, which hinders understanding of how their populations differentiate in response to environmental drivers or host organisms. Their mechanisms employed to cope with oxidative stress are largely unknown. Here, we utilized genome-resolved metagenomics to investigate the adaptive genome repertoire of Patescibacteria in oxic and anoxic groundwaters, and to infer putative host ranges.Results: Within six groundwater wells,Cand. Patescibacteria was the most dominant (up to 79%) super-phylum across 32 metagenomes sequenced from DNA retained on 0.2 and 0.1 µm filters after sequential filtration. Of the reconstructed 1275 metagenome-assembled genomes (MAGs), 291 high-quality MAGs were classified asCand. Patescibacteria.Cand. Paceibacteria andCand. Microgenomates were enriched exclusively in the 0.1 µm fractions, whereas candidate division ABY1 andCand. Gracilibacteria were enriched in the 0.2 µm fractions. On average, Patescibacteria enriched in the smaller 0.1 µm filter fractions had 22% smaller genomes, 13.4% lower replication measures, higher proportion of rod-shape determining proteins, and of genomic features suggesting type IV pili mediated cell–cell attachments. Near-surface wells harbored Patescibacteria with higher replication rates than anoxic downstream wells characterized by longer water residence time. Except prevalence of superoxide dismutase genes in Patescibacteria MAGs enriched in oxic groundwaters (83%), no major metabolic or phylogenetic differences were observed. The most abundant Patescibacteria MAG in oxic groundwater encoded a nitrate transporter, nitrite reductase, and F-type ATPase, suggesting an alternative energy conservation mechanism. Patescibacteria consistently co-occurred with one another or with members of phyla Nanoarchaeota, Bacteroidota, Nitrospirota, and Omnitrophota. Among the MAGs enriched in 0.2 µm fractions,, only 8% Patescibacteria showed highly significant one-to-one correlation, mostly with Omnitrophota. Motility and transport related genes in certain Patescibacteria were highly similar to genes from other phyla (Omnitrophota, Proteobacteria and Nanoarchaeota).Conclusion: Other than genes to cope with oxidative stress, we found little genomic evidence for niche adaptation of Patescibacteria to oxic or anoxic groundwaters. Given that we could detect specific host preference only for a few MAGs, we speculate that the majority of Patescibacteria is able to attach multiple hosts just long enough to loot or exchange supplies.

Taxonomic classification, that is, the assignment to biological clades with shared ancestry, is a common task in genetics, mainly based on a genome similarity search of large genome databases. The classification quality depends heavily on the database, since representative relatives must be present. Many genomic sequences cannot be classified at all or only with a high misclassification rate. Here we present BERTax, a deep neural network program based on natural language processing to precisely classify the superkingdom and phylum of DNA sequences taxonomically without the need for a known representative relative from a database. We show BERTax to be at least on par with the state-of-the-art approaches when taxonomically similar species are part of the training data. For novel organisms, however, BERTax clearly outperforms any existing approach. Finally, we show that BERTax can also be combined with database approaches to further increase the prediction quality in almost all cases. Since BERTax is not based on similar entries in databases, it allows precise taxonomic classification of a broader range of genomic sequences, thus increasing the overall information gain.