Genome-wide multi-layer annotation of RNA regulatory features in human endogenous retroviruses across the GRCh38 genome

💡Introduction This dataset provides a comprehensive, genome-wide annotation of RNA regulatory features embedded within human endogenous retroviruses (HERVs) in the human reference genome (GRCh38). Building on previously defined HERV internal regions and long terminal repeats (LTR) annotations, we systematically characterized RNA-centered regulatory potential across 53,176 internal regions and associated LTRs. Using a reproducible computational pipeline based on motif scanning, genomic overlap analysis, and transcript annotation integration, we identified widespread RNA-binding protein binding sites (RBPBS), revealing structured and subfamily-specific regulatory landscapes. In total, the vast majority of HERV internal regions harbor potential RBP interactions, with distinct enrichment profiles observed across major HERV families. To place these features in biological context, we integrated HERV annotations with transcript architectures, identifying thousands of long non-coding RNAs (lncRNAs) and protein-coding genes containing embedded HERV sequences. Conserved retroviral domains were further mapped to transcript features, revealing a strong enrichment in terminal exons and 3′ untranslated regions (3′ UTRs), consistent with RNA-level regulatory and potential micropeptide-coding roles. Functional annotation of HERV-associated lncRNAs was performed through integration of curated databases and large-scale literature mining, highlighting a largely unexplored class of transcripts enriched in retroviral sequence. In addition, we identified more than 6,500 antisense LTR insertions located in transcript termini, representing candidate SPARCS-like configurations with the potential to generate dsRNA and contribute to innate immune signaling. 🧩 Integration into HERVarium This dataset constitutes the RNA regulatory layer of HERVarium (https://hervarium.cnag.eu/), a comprehensive reference resource for HERVs that integrates internal domain annotation and LTR regulatory features. HERVarium aims to provide a unified, multi-layer framework for exploring HERV structure, coding potential, and regulatory architecture at both the DNA and RNA levels. This dataset complements previously released resources describing conserved retroviral protein domains and LTR transcription factor binding landscapes. Together, these datasets enable integrative exploration of HERV biology, spanning sequence conservation, transcript architecture, and post-transcriptional regulatory potential. 🧬 The dataset includes Genome-wide predictions of RNA-binding protein binding sites (RBPBS) across HERV internal regions Integration of HERV sequences within transcript architectures, including lncRNAs and protein-coding genes Mapping of conserved retroviral domains to transcript features (exons, CDS, UTRs) Identification of antisense LTR configurations compatible with dsRNA formation (SPARCS-like candidates) Functional annotation of HERV-associated lncRNAs using curated databases and literature mining Characterization of interferon-related regulatory potential in LTRs via STAT and IRF motif analysis Together, these resources provide a multi-layer view of RNA regulatory features encoded by HERV sequences and support future research into post-transcriptional regulation, immune signaling, and evolutionary co-option of transposable elements. 📄 Description of each file domain_tx_feature_overlaps.tsv: overlap of conserved HERV-derived protein domains with transcript features (exons, CDS, UTRs), including positional and strand-aware annotations. herv_lncrna_with_literature.tsv: integrated table of lncRNAs overlapping HERV internal regions with functional annotations derived from databases and large-scale literature mining. ltr_candidates_antisense_last_exon_or_utr3.protein_coding_or_lncRNA.tsv: catalogue of antisense LTR insertions located in transcript termini (3′ UTRs or last exons), representing candidate SPARCS-like loci. LTR_IFN_STAT1STAT2_IRF_summary.tsv: summary of interferon-related regulatory potential in LTRs based on co-localized STAT1/STAT2 and IRF motif architectures. LTR_IFN_STAT1_summary.tsv: summary of STAT1 motif occurrence and distribution across HERV LTR sequences. RBP_fimo.tsv: genome-wide predictions of RNA-binding protein binding sites within HERV internal regions obtained by FIMO motif scanning. RBP_fimo.bed: BED file with genomic coordinates of predicted RNA-binding protein binding sites (RBPBS) in HERV internal regions derived from FIMO motif scanning. SPARCS_like_candidates_with_U3R_U5_catalogue.tsv: annotated SPARCS-like LTR candidates integrated with U3–R–U5 structural features and reconstruction metrics. transcript_context.with_herv_id.tsv: comprehensive mapping of HERV internal regions to genomic and transcript contexts, including gene overlap, strand orientation, and locus identifiers. ATtRACT_db.txt: tab-separated database of RNA-binding protein motifs from ATtRACT, including gene names, motif sequences, experimental evidence, and matrix identifiers used for downstream motif analysis. pwm.txt: position weight matrix (PWM) file in MEME-compatible format containing RNA-binding protein motifs used for FIMO scanning across HERV internal regions. hervs_internal_and_ltrs.gtf: GTF annotation of HERV loci integrating internal regions and LTRs into unified transcriptional units, enabling downstream transcriptomic and regulatory analyses. lncRNA_herv_coverage.tsv: quantification of HERV internal sequence contribution to human lncRNA gene bodies and exons, including overlap metrics, fractions, and associated HERV subfamilies. protein_coding_overlapping_herv_coverage.tsv: quantification of HERV internal sequence overlap in human protein-coding genes, reporting gene-level and exonic coverage, fractions, and associated HERV subfamilies. HERV_domains_transcript_context_last_exon.xlsx: genome-wide annotation of HERV-derived protein domains overlapping terminal exons of human transcripts (GENCODE v48), including both protein-coding genes and lncRNAs. The dataset reports domain type, coverage, subfamily, and transcript context features. 💻 Code and citation The dataset was generated using a custom computational pipeline integrating motif scanning (FIMO), genomic interval analysis (BEDTools), and transcript annotation (GENCODE), which is available at: Montserrat-Ayuso et al., Human endogenous retroviruses shape the RNA regulatory landscape of the human genome 🔗 https://github.com/funcgen/herv-rna-regulation If you use this dataset in your research, please cite both the dataset DOI and the corresponding manuscript: HERVs as building blocks of RNA regulatory architecture in the human genome Tomàs Montserrat-Ayuso, Aurora Pujol, Anna Esteve-Codina bioRxiv 2026.04.29.721355; doi: https://doi.org/10.64898/2026.04.29.721355 Version history v1: Initial release of genome-wide RNA regulatory annotations across HERV internal regions and LTRs.