You carry two copies of every chromosome, one from each parent, and those copies can differ in ways that matter enormously for disease risk and drug response. Assembling both copies end to end, and determining which variant came from which parent, is the true goal of personalized genomics. Today's leading approaches each fall short in different ways: long reads phase locally but struggle to span entire chromosomes; Hi-C requires deep sequencing (>30x) for accurate phasing; optical mapping lacks sequence-level resolution. None delivers the complete, phased picture that personalized genomics demands.
Into this gap steps Strand-seq, a technology first described in 2012 that, thanks to advances in automation, single cell workflows, and sequencing technology, is today proving indispensable for the most ambitious genome analyses in the world.
Developed by Dr. Peter Lansdorp's laboratory at the BC Cancer Research Institute in Vancouver and introduced in a landmark 2012 Nature Methods paper, Strand-seq selectively sequences only the original template strands from a single cell, preserving their directionality. Cells are cultured with bromodeoxyuridine (BrdU) for one round of DNA replication, labeling newly synthesized strands while leaving the original templates intact. After FACS sorting of single nuclei, BrdU-containing strands in DNA fragments ligated to forked adapters are nicked by UV light before PCR, so only the original templates are amplified. Every resulting read maps to either the forward (Crick) or reverse (Watson) strand of the reference genome, directly encoded by the parental template strand it came from.
By examining these Watson/Crick patterns across dozens of cells, Strand-seq enables chromosome-length haplotyping, detection of large structural variants such as inversions, genome assembly scaffolding, and more, at far lower sequencing cost than competing methods.
A critical contribution to making this data interpretable has come from bioinformaticians including Dr. David Porubsky, who developed breakpointR, an R/Bioconductor package that detects strand-state changes with a read-based sliding window approach, outperforming earlier fixed-bin methods by 10–30% in sensitivity.
One of Strand-seq's most distinctive capabilities is the high-resolution mapping of sister chromatid exchange events (SCEs), something essentially impossible with other sequencing approaches. SCEs are markers of genomic stress arising from double-strand break repair, and their accumulation is a hallmark of Bloom syndrome, a rare cancer-prone genetic instability disorder. Strand-seq resolves SCEs at orders of magnitude greater resolution than cytogenetic banding, has provided new mechanistic insights into Bloom syndrome, and holds promise as a sensitive genomic instability assay across oncology and beyond.
Strand-seq remains in routine use in only a handful of labs due to its technical demands: live cells for BrdU incorporation, specialized FACS sorting, and nanoliter-scale library preparation from picogram quantities of DNA. Dr. Lansdorp's lab has worked steadily to lower these barriers through improved protocols, automation, and computational tools.
As Strand-seq scaled to applications requiring higher total sequencing output for hundreds to thousands of pooled single cell libraries, an unexpected problem emerged. When leading sequencing platforms moved to patterned flow cells, the lab began seeing noise in strand assignment data from those instruments caused by index hopping, an artifact of index mis-assignment between reads of multiplexed libraries that is heightened by the presence of residual contaminating free adapters. For most applications, a few percent noise in sequence data is hard to detect, but inaccurate read assignment interferes with downstream analysis. Index hopping is readily detected with Strand-seq as background reads that reflect incorrect strand-directional read-to-genome alignment. The standard mitigation strategy of unique dual indexes does not scale to the thousands of libraries required for high-throughput Strand-seq data production.
The solution came from the Element Biosciences AVITI™ platform, which uses rolling circle amplification (RCA). Because each DNA molecule amplifies in place from a circularized template, there is no mechanism for index hopping. The improvement was unambiguous: mis-indexing noise dropped to near zero and strand assignment accuracy was restored to near-perfect levels, while also delivering the throughput and cost-effectiveness that large Strand-seq experiments demand.
With tools like breakpointR making Strand-seq data interpretable at scale, and platforms like AVITI resolving the sequencing challenges of high-throughput library pooling, the barriers keeping this technology in specialist hands are gradually coming down. The method that mapped SCEs at a resolution of a few kilobases or less in 2012, combined with long reads, is today enabling fully phased diploid human genome assemblies. Its role in personalized genomics is only just beginning.
Falconer, E., Hills, M., Naumann, U. et al. DNA template strand sequencing of single-cells maps genomic rearrangements at high resolution. Nat Methods 9, 1107–1112 (2012). https://doi.org/10.1038/nmeth.2206
Hanlon, V.C.T., Lansdorp, P.M. Strand-seq and the future of personalized genomics. Nat Genet (2026). https://doi.org/10.1038/s41588-026-02548-4
David Porubsky, Ashley D Sanders, Aaron Taudt, Maria Colomé-Tatché, Peter M Lansdorp, Victor Guryev, breakpointR: an R/Bioconductor package to localize strand state changes in Strand-seq data, Bioinformatics, Volume 36, Issue 4, February 2020, Pages 1260–1261, https://doi.org/10.1093/bioinformatics/btz681