Sixteen diverse laboratory mouse reference genomes define strain-specific haplotypes and novel functional loci

Jingtao Lilue, Anthony Doran, Ian T. Fiddes, Monica Abrudan, Joel Armstrong, Ruth Bennett, William Chow, Joanna Collins, Stephan C. Collins, Anne Czechanski, Petr Danecek, Mark Diekhans, Dirk-Dominik Dolle, Matt Dunn, Richard Durbin, Dent Earl, Anne C. Ferguson‐Smith, Paul Flicek, Jonathan Flint, Adam Frankish, Beiyuan Fu, Mark Gerstein, James Gilbert, Leo Goodstadt, Jennifer Harrow, Kerstin Howe, Ximena Ibarra-Soria, Mikhail Kolmogorov, Chris Lelliott, Darren W. Logan, Jane Loveland, Clayton E. Mathews, Richard Mott, Paul Muir, Stefanie Nachtweide, Fábio C. P. Navarro, Duncan T. Odom, Naomi Park, Sarah Pelan, Son Pham, Mike Quail, Laura G. Reinholdt, Lars Romoth, Lesley Shirley, Cristina Sisu, Marcela Sjöberg, Mario Stanke, Charles A. Steward, Mark Thomas, Glen Threadgold, David Thybert, James Torrance, Kim Wong, Jonathan Wood, Binnaz Yalcin, Fengtang Yang, David J. Adams, Benedict Paten, Thomas KeaneNature Genetics 2018 · DOI

Introduction

This research, published in Nature Genetics, represents a major leap forward in mammalian genomics by moving beyond the "single reference genome" paradigm. By performing de novo assemblies of 16 diverse inbred mouse strains, the authors have mapped the "dark matter" of the mouse genome—highly divergent, strain-specific regions that were previously invisible to standard analysis.

The Problem: Overcoming Reference Bias

For nearly two decades, mouse genetics has relied almost exclusively on a single reference genome: the C57BL/6J strain (the "Black 6" mouse). While revolutionary, this approach has a major flaw: when researchers sequence other strains, they typically "map" the new data back to this single reference.

If a different strain has a large chunk of DNA that C57BL/6J lacks, or if a region is so different that the sequences don't align, that variation is effectively lost. This "reference bias" hides the very diversity that makes different mouse strains useful for modeling human diseases. The authors sought to eliminate this bias by building complete, "ground-up" (de novo) assemblies for 16 of the most important laboratory mouse strains.

Methods: A Multi-Platform Assembly Pipeline

The team employed a sophisticated multi-platform assembly pipeline:

  • Sequencing: Illumina short-read sequencing (40–70× coverage) combined with Dovetail Chicago libraries for long-range linkage.
  • Assembly & Scaffolding: Initial contigs built with the SGA assembler were organized into chromosome-scale "pseudochromosomes" using Ragout.
  • Annotation: Tools like Comparative Augustus and transMap were used to identify genes de novo, incorporating strain-specific RNA-Seq data.
  • Validation: CRISPR/Cas9 was used to create knockout mice to verify the function of newly discovered loci.

Datasets: Capturing Global Mouse Diversity

The study utilized 16 inbred mouse strains, including 12 classical strains (e.g., A/J, BALB/cJ, DBA/2J, NOD/ShiLtJ) and 4 wild-derived strains (CAST/EiJ, PWK/PhJ, WSB/EiJ, and SPRET/EiJ). These wild-derived strains represent significantly higher genetic diversity than classical strains.

Key Insights: Hotspots of Diversity

The authors identified 2,567 regions of "extreme allelic variation." These hotspots are heavily enriched for genes involved in pathogen defense, immunity, and reproduction, suggesting that evolutionary pressure from viruses and bacteria is a primary driver of genomic innovation.

Strain-specific haplotypes and high-diversity regions on Chromosome 11, including the Nlrp1 immune locus.

The Engine of Variation

These high-diversity regions are packed with transposable elements (retrotransposons like LINEs and LTRs). These "jumping genes" appear to be the mechanism for rapid genomic shuffling, allowing gene families to expand or contract between strains.

Analysis of transposable element age and divergence within and outside high-diversity regions.

Key Results: Functional Discovery and Validation

  • Improved Reference: The study filled gaps in the current mouse reference (GRCm38), completing 10 gene structures and adding 62 new coding loci.
  • Structural Discovery: They identified 23 distinct clusters of olfactory receptors and 7 distinct versions of the Nlrp1 immune locus across the strains.
  • Functional Validation: Knocking out the newly discovered Efcab3-like gene—a massive gene previously unannotated due to its complexity—resulted in severe brain developmental anomalies, including a 65% enlargement of the lateral ventricles and increased size of the cerebellum.

Limitations and Future Directions

While these are chromosome-scale assemblies, they remain draft genomes with gaps in highly repetitive centromeric regions. Future work will likely involve using longer-read technologies (like Oxford Nanopore or PacBio HiFi) and expanding the effort to hundreds of other specialized mouse strains to fully capture the functional landscape of the Mus genus.

Summary updated

BibTeX
@inproceedings{lilue2018sixteen,
  author = {Jingtao Lilue and Anthony Doran and Ian T. Fiddes and Monica Abrudan and Joel Armstrong and Ruth Bennett and William Chow and Joanna Collins and Stephan C. Collins and Anne Czechanski and Petr Danecek and Mark Diekhans and Dirk-Dominik Dolle and Matt Dunn and Richard Durbin and Dent Earl and Anne C. Ferguson‐Smith and Paul Flicek and Jonathan Flint and Adam Frankish and Beiyuan Fu and Mark Gerstein and James Gilbert and Leo Goodstadt and Jennifer Harrow and Kerstin Howe and Ximena Ibarra-Soria and Mikhail Kolmogorov and Chris Lelliott and Darren W. Logan and Jane Loveland and Clayton E. Mathews and Richard Mott and Paul Muir and Stefanie Nachtweide and Fábio C. P. Navarro and Duncan T. Odom and Naomi Park and Sarah Pelan and Son Pham and Mike Quail and Laura G. Reinholdt and Lars Romoth and Lesley Shirley and Cristina Sisu and Marcela Sjöberg and Mario Stanke and Charles A. Steward and Mark Thomas and Glen Threadgold and David Thybert and James Torrance and Kim Wong and Jonathan Wood and Binnaz Yalcin and Fengtang Yang and David J. Adams and Benedict Paten and Thomas Keane},
  title = {{Sixteen diverse laboratory mouse reference genomes define strain-specific haplotypes and novel functional loci}},
  booktitle = {Nature Genetics},
  year = {2018},
  doi = {10.1038/s41588-018-0223-8},
}
Figure