Yun Deng

Homepage - 2026 Population, Evolutionary, and Quantitative Genetics Conference Visit our website to learn more.

PEQG website live! genetics-gsa.org/peqg-2026/

Somatic mutation and selection at population scale - Nature A new version of nanorate DNA sequencing, with an error rate lower than five errors per billion base pairs and compatible with whole-exome and targeted capture, enables epidemiological-scale studies of somatic mutation and selection and the generation of high-resolution selection maps across coding and non-coding sites for many genes.

Our latest work is out in Nature today. In this paper, we introduce an improved version of NanoSeq, a duplex sequencing protocol with <5 errors per billion bp in single DNA molecules, and use it to study the somatic mutation landscape of oral epithelium in >1000 people www.nature.com/articles/s41...

On ARGs, pedigrees, and genetic relatedness matrices Abstract. Genetic relatedness is a central concept in genetics, underpinning studies of population and quantitative genetics in human, animal, and plant se

Our paper (Brieuc Lehmann is the first author) is now online in Genetics.

academic.oup.com/genetics/adv...

Sperm sequencing reveals extensive positive selection in the male germline - Nature A combination of whole-genome NanoSeq with deep whole-exome and targeted NanoSeq is used to accurately characterize mutation rates and genes under positive selection in sperm cells.

Now published! Our paper on:
(1) Accurate sequencing of sperm at scale
(2) Positive selection of spermatogenesis driver mutations across the exome
(3) Offspring disease risks from male reproductive aging
[1/n]
www.nature.com/articles/s41...

Inference of complex demographic history using composite likelihood based on whole-genome genealogies Accurate parametric inference on complex demographic models is a continuing challenge in population genetics. Ancestral recombination graphs (ARGs) provide richer information than simple population ge...

Excited to preprint our latest work (w/ Drew DeHaas, Zhibai Jia, Leo Speidel) on using ARGs for demographic inference. w/ applications using data from 1000 Genomes Project. www.biorxiv.org/content/10.1...

I want to try something again at #ASHG25 this year: I'll block some time on Thursday and Friday afternoons to meet with trainees who would be interested to chat on any topic.

I did this last year and it was great to meet a whole bunch of new people, at all career stages!

The Length of Haplotype Blocks and Signals of Structural Variation in Reconstructed Genealogies Abstract. Recent breakthroughs have enabled the accurate inference of large-scale genealogies. Through modelling the impact of recombination on the correla

Delighted that our paper about the distribution of genomic spans of clades/edges in genealogies (ARGs), and using this for detecting inversions and other SVs (and other phenomena that cause local disruption of recombination) is out in MBE academic.oup.com/mbe/article/... (1/n)

We are excited to share GPN-Star, a cost-effective, biologically grounded genomic language modeling framework that achieves state-of-the-art performance across a wide range of variant effect prediction tasks relevant to human genetics.
www.biorxiv.org/content/10.1...
(1/n)

Btw do you have any insights about the acceleration performance from SqS3 versus quasi-Newton updates? Is one of them significantly better than the other, or do they perform roughly equally well? Thanks!

We did more comprehensive benchmark and expanded the empirical analysis for the first round, not so much for the following rounds. The methodology never changes throughout the revisions.

SINGER is a fantastic method for ARG Inference and based on some of the analyses I've seen definitely the best trade off between accuracy and sample size

Robust and accurate Bayesian inference of genome-wide genealogies for hundreds of genomes - Nature Genetics SINGER is a method for creating ancestral recombination graphs to understand the genealogical history of genomes. The method has increased speed, and thus scalability, without sacrificing accuracy.

SINGER, our ARG inference method, is finally published and freely available online:

doi.org/10.1038/s415...

It was a long journey – 16 months from initial submission to acceptance. Is it just me, or has peer review gotten more arduous lately? 4+ rounds of review isn't so unusual these days...

Super useful insights! Thanks a lot!

Hi Arun, thanks for sharing this work! Any particular reason you performed acceleration on the EM updates rather than the quasi-Newton updates as ADMIXTURE does? Is it because the convergence is already fast with accelerated EM?