Significant diversity of human anelloviruses revealed by novel viral sequences identified in human metagenomic data

Human torque teno viruses are emerging infectious agents distributed globally and have increasingly been reported to be associated with human diseases. To identify potential anelloviral sequences in available metagenomic data, an in silico screening was performed mainly employing the ORF1, ORF2 and ORF3 nucleotide/protein queries of known human anelloviruses and identified 217 complete ORF1 regions. Pairwise nucleotide‐identity analysis with a 69% cut‐off – consistent with ICTV species demarcation – revealed 117 novel species across the 3 major human‐infecting genera: 15 in Alphatorquevirus , 51 in Betatorquevirus and 51 in Gammatorquevirus . In nearly all cases, these species assignments correspond precisely to monophyletic clusters in maximum‐likelihood phylogenies of ORF1 amino acid sequences. Using AlphaFold3-guided modelling together with representative ORF1 alignments, we delineated capsid motifs – the conserved jelly-roll (JR) β -sandwich core ( β -strands B–I) and the outward projection domains P1/P2 – and quantified motif lengths across genera, revealing tightly constrained JR lengths with genus-specific but overlapping variation in P1/P2. A few exceptions – where pairwise‐based groupings split or merge slightly differently – highlight ongoing challenges in delineating rapidly evolving viruses. Notably, the two deeply branching isolates retain the canonical JR core while exhibiting a TTMDV-like short P2, indicating preservation of key capsid architecture in the newly proposed genus. This work nearly doubles the known species richness of human anelloviruses and introduces a novel genus, underscoring the vast, hidden diversity of the gut virome and its potential impact on human health. By coupling taxonomy with structure-informed ORF1 motif analysis, our study provides biological context for these lineages and a framework for future functional and immunological investigations.