henhaobio
Sign in

Journal of Evolutionary BiologyGenomicsPeer-reviewed journal article

ntSynt-viz: Visualizing synteny patterns across multiple genomes

Lauren CoombeiD, René L. WarreniD, İnanç BiroliD

PDFDOIPreprintCodeData
In-depth full-text digest available · Jump to digest ↓

Abstract

With the explosion of chromosome-scale genome assemblies being generated in recent years, there is vast potential for comparative genomics analyses through detecting multi-genome synteny. While existing tools can detect synteny blocks between multiple genomes, their text-based outputs make it challenging to intuitively explore large-scale synteny patterns. Interpretable, information-rich and easy-to-use synteny visualization tools are imperative to enable important biological insights from the synteny block data output by the aforementioned utilities. Here, we present ntSynt-viz, a command-line tool for automated sorting, normalization and plotting of multi-genome synteny blocks. We show how ntSynt-viz provides clearer and more easily interpretable chromosome-painting ribbon plots compared to the state-of-the-art tool NGenomeSyn when evaluating synteny between 14 human genomes and 9 hoverfly genomes. We expect that ntSynt-viz will provide crucial insights into large-scale synteny patterns between divergent genomes, thereby advancing research into key evolutionary questions. ntSynt-viz is freely available on GitHub (https://github.com/bcgsc/ntsynt-viz).

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (3 Feb 2025) · CC BY 4.0

ntSynt-viz sorts genomes and chromosomes by synteny, normalizes strands, and draws chromosome-painting ribbon plots easier to interpret than NGenomeSyn output for human and hoverfly genomes.

Multi-genome synteny detectors such as ntSynt output structured text that is hard to read, while existing ribbon plotters (NGenomeSyn, plotsr) and R packages (gggenomes, syntenyPlotteR) lack built-in synteny-block pre-processing and the latter need heavy input reformatting. ntSynt-viz wraps synteny-driven genome ordering, chromosome ordering and strand normalization into one command so rearrangements across many genomes become visible without manual curation.

Key results

  1. SuggestedOrdering genomes and chromosomes by synteny across 9 Cheilosia hoverfly assemblies revealed a broadly conserved chromosome and a fusion shared by C. scutellata and C. soror, the most rearranged genomes in the set.

    Assemblies span 5–7 chromosomes and 354–545 Mbp. Content of OX411880.1 (C. impressa) is retained in all compared genomes despite inversions. The shared fusion joins segments of 2 C. urbana chromosomes and 3 chromosomes in the other genomes. On the OX411880.1-syntenic chromosomes, most gray (unsyntenic) regions lie near the centre, and genome sizes differ by up to ∼200 Mbp. · Results and Discussion; Figure legends

  2. SuggestedNGenomeSyn plots of the same hoverfly synteny blocks were harder to interpret because chromosomes and genomes stay in arbitrary order. Producing them also needed manual scripting rather than a single command.

    Visual comparison of Figure 3 (NGenomeSyn v1.41) with Figure 2 (ntSynt-viz v1.0.0): conserved OX411880.1 is not readily seen, crossing ribbons can come from arbitrary chromosome order, and the C. scutellata/C. soror fusion is hard to discern. NGenomeSyn needed manually customized configuration files. No quantitative interpretability metric is reported. · Testing ntSynt-viz and NGenomeSyn; Results and Discussion; Figure legends

  3. ShownStrand normalization against the target genome reverse-complemented chromosomes that had been assembled on the opposite strand. This removed technical orientation artifacts and made collinearity and inversions easier to read in both datasets.

    Human: 1, 12 and 9 chromosomes were reverse-complemented in KSA001 mat, PGP1 and KOREF relative to T2T (compare with NGenomeSyn, Supplementary Fig S3). Hoverfly: normalizing the C. vulpina chromosome syntenic to OX411881.1 (C. impressa) exposed large inversions vs C. variabilis and collinearity with C. vernalis, which were unclear without normalization (Supplementary Figure S4). · Strand normalization; Results and Discussion; Figure legends

  4. SuggestedZooming into human chromosome 8 with --keep showed an inversion near the start of the chromosome, carried by the T2T, HG01243 and HG002 assemblies. It corresponds to a known polymorphic human inversion.

    14 autosome-only assemblies from 10 individuals. Twisted ribbons appear between adjacent pairs (HG01243/CN1 mat, KSA001 mat/HG002 pat, HG002 mat/HuRef alt), marking which genomes carry the inverted orientation. The authors match the region to a known polymorphic inversion; no orthogonal validation is shown. · Results and Discussion; Figure legends

  5. ShownntSynt-viz renders plots more slowly than NGenomeSyn, but both tools finish within minutes.

    Visualization time was 1.0 min for ntSynt-viz vs 0.1 min for NGenomeSyn; both took under 2 minutes across all runs (Supplementary Table S5). Hardware is not specified. · Results and Discussion

Methods, data and reproducibility

Snakemake-driven Python CLI that scores pairwise synteny distances from adjacent-block consistency, builds a Quicktree neighbour-joining cladogram to order genomes, orders chromosomes by tile best-hit mapping between consecutive genomes (default 1 Mbp tiles), optionally reverse-complements chromosomes to a target genome, and plots with gggenomes/ggtree.

  • Chromosome-scale human genome assemblies, autosomes only, incl. T2T (target), HG002, HG01243, PGP1, KOREF, KSA001, CN1, HuRef; ntSynt run with -d 0.5 · human · n = 14 assemblies from 10 individuals (5 with diploid assemblies)
  • Cheilosia hoverfly genome assemblies from the Earth BioGenome Project database, sequenced by Darwin Tree of Life; 5–7 chromosomes, 354–545 Mbp; ntSynt run with -d 20 · Cheilosia spp. (hoverflies) · n = 9 genomes · Supplementary Table S3; chromosome IDs cited include OX411880.1, OX411881.1, OX411884.1 (C. impressa)
  • Annotated T2T centromere coordinates, drawn as black bars via --centromeres · human
  • NGenomeSyn configuration files used to make the comparison plots · 10.5281/zenodo.14247722
Methods & tools
ntSynt v1.0.2 (multi-genome synteny block detection), ntSynt-viz v1.0.0, NGenomeSyn v1.41 (comparator), Quicktree (neighbour-joining cladogram), gggenomes (ribbon plot rendering), ggtree (cladogram plotting), Snakemake (pipeline orchestration), SyRI and halSynteny (supported via format conversion scripts)
Software & environment
ntSynt v1.0.2 (defaults except -d 0.5 human, -d 20 hoverfly), ntSynt-viz v1.0.0 (Snakemake-driven), NGenomeSyn v1.41; exact commands in Supplementary Table S4; no container or dependency versions stated.
Assay conditions & compute
Visualization took 1.0 min (ntSynt-viz) vs 0.1 min (NGenomeSyn), with all runs under 2 minutes (Supplementary Table S5); hardware, CPU count and memory not stated.
Reproducibility gaps
Assembly accessions (Supplementary Tables S2, S3), run commands (Table S4) and parameter documentation (Table S1) appear only in supplementary material; the text gives no software licence or container specification.
Limitations (4)
  • NGenomeSyn rendered plots faster than ntSynt-viz (0.1 min vs. 1.0 min), although the authors note both finished in under 2 minutes. · stated by the authors
  • The authors note that plotting many small sequences from fragmented assemblies can obscure overall synteny patterns, and suggest optional filtering by sequence length (--seq_length) and block length (--length) for clearer plots. · stated by the authors
  • Interpretability was judged only by visual comparison against a single comparator (NGenomeSyn). plotsr, gggenomes and syntenyPlotteR were not run, and there was no user study or quantitative readability metric. · noted in this summary
  • Only two chromosome-scale datasets were shown. The tool was not tested on fragmented assemblies, larger genome counts, or blocks converted from other tools, so robustness of the synteny-distance genome ordering and tile-based chromosome ordering there is unknown. · noted in this summary

Read this if you compare several chromosome-scale assemblies and want single-command ribbon plots that auto-order genomes and chromosomes and normalize strands from ntSynt or converted SyRI/halSynteny blocks.

Citations
23
FWCI
44.04
Field percentile
top 1%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISCoombe '25Gibbons '26Warren '18Julian '24Istace '26Almeida‐Sil '22
75% semantic matchBioinformatics
Lauren Coombe, Parham Kazemi, Johnathan Wong+2 more·BMC Biology·Dec 29, 2025· 11 citations