RESEARCH INTERESTS

Variations are the raw material of evolution. How, where, and why they arise, and how often they are introduced into the genome, are fundamental questions, not only for evolutionary biology but in the study of genetic disease. Specific research areas are:

Spermatogenesis, germline mutation, and the testis as a ‘crucible of creation’

De novo mutations make a significant contribution to human disease, affecting approximately 1 in 300 new births. The vast majority of these (>80%) have their origins in the paternal germline, having occurred as spontaneous copy errors during spermatogenesis (the process of sperm production). This entails spermatogonial stem cells (SSCs) and their progenies regularly dividing throughout a reproductive lifespan, and (to ensure reproductive success) to do so with minimal error. Yet, despite this constant – and intrinsically error-prone – cellular activity, germline mutation rates are lower than those of somatic cells in all species studied to date. This suggests that the mechanisms by which spermatogenesis is regulated play a fundamental role in fine-tuning the germline mutation rate, the pace at which new mutations – the raw material of evolution – are introduced into the genome.

The testis is therefore something of a ‘crucible’ in which DNMs form and propagate across generations, yet spermatogenesis has been under-studied in humans: it is challenging to procure testicular samples. Nevertheless, the field has been galvanised in recent years by advances in single-cell sequencing. I am currently using single-cell data to study the mechanisms of sperm production and the regulation of SSCs, as this can give insight into the factors controlling germline mutation rate.

In addition, although the majority of DNMs are one-off events, they can also come about by the process of gonadal mosaicism. This requires that the DNM occurred at a very early developmental stage for one of the parents, and consequently becomes restricted to the germline. This means that – resembling a mosaic – they are found in multiple gonadal cells (eggs or sperm) but no other cell of the body. Distinguishing DNMs which arose from gonadal mosaicism from the more common ‘one-offs’ has important clinical applications. Following the birth of a child with a disorder caused by a DNM, parents often ask ‘what is the risk that another future child of ours will be affected?’ I was previously involved in the PREGCARE (PREcision Genetic Counselling And Reproduction) study, which establishes a systematic, evidence-based, approach to providing personalised recurrence risk assessment. This will allow couples to make informed choices about the different diagnostic options available to them during future pregnancies.

Alternative splicing and the ‘landscape of opportunities’

The vast majority of genes in multicellular eukaryotes generate multiple transcripts: alternative isoforms. As alternative splicing can increase the number of biological functions encoded in a genome without increasing the number of genes, it has played a prominent role in many major innovations over evolutionary time (such as brain development in animals). For instance, work undertaken during my PhD demonstrated that alternative splicing is the foremost predictor of organism complexity and that it is associated with phenotypic novelty. More recently, I have been involved in cataloguing metazoan splice variants throughout development.

However, while a number of alternative transcripts are selectively beneficial, the vast majority appear to be non-functional (being low-level noise in the splicing machinery) and so alternative splicing, as a whole, is not generally adaptive. Nevertheless, this ‘transcriptional exhaust’ may in itself provide the substrate for selection by creating a ‘landscape of opportunities’ from which functional variants may arise. An ongoing research interest is in charting this landscape. I am especially interested in the role alternative splicing plays in the testis, as this is not only among the most transcriptionally rich tissues in all organisms studied to date, but because alternative isoforms are known to increase across the length of the spermatogenic trajectory (ostensibly because the ‘opportunities’ which arise may increase the likelihood of reproductive success).

Variant calling & its benchmarking

Given the central role of DNA and RNA variants as ‘raw material’, it is critical to call them accurately. Comparative evaluation of methodologies is also necessary to stay abreast of algorithmic advances and in refining best practices and to that end I have an interest in the development and use of benchmarking datasets. In the context of microbial genomics, I have made comparative evaluations of tools and strategies for SNP calling and filtering, as well as de-contaminating and trimming bacterial reads. More recently, I have been involved in the development of the Chinese Quartet benchmarking resources, as well as methodologies for calling particularly complex variant categories (namely, large indels and structural variants), such as SVision-pro.

Onomastics (names and naming)

Although I am a computational biologist I have had a personal and academic interest in names and naming for many years. In my onomastic research I apply the tools and techniques of bioinformatics to quantitative name data, centred at present largely on the UK BMD (birth, marriage, death) registers. As names are a product and reflection of cultural changes over time, I am ultimately interested in understanding how these came about, what factors influenced them, and how they spread.

These questions of cultural evolution complement those of biological evolution. How are cultural variants – styles, ideas, practices, and tastes – transmitted (copied between people) and what mechanisms govern their success? To explore this question, I use baby names as a unit of analysis – they not only classify a person as an individual but as a product and reflection of society. I am currently interested in assembling large-scale name use datasets such that transmission biases which predispose someone to favour one cultural variant over another may be identified; these can give insight into the broader mechanisms of cultural evolution.

A subsidiary interest is in adapting bioinformatic methods to this non-biological data. For example, I have applied network analysis methods previously used for gene expression data to almost two centuries of UK birth records. This demonstrates how social and demographic changes manifest, by analogy to genes, in the ‘co-expression’ of personal names. A website was created to visualise this which received media coverage from the BBC and The Guardian, and an interview with the Dutch paper Trouw.

CAREER OVERVIEW

2010-2014

PhD evolutionary genomics, University of Bath.

进化基因组学博士巴斯大学.

2014-2017

Postdoctoral research fellow, The Roslin Institute, University of Edinburgh.

博士后研究员爱丁堡大学罗斯林研究所.

2018-2023

Senior bioinformatician, University of Oxford (Nuffield Department of Medicine 2018-2021, Weatherall Institute of Molecular Medicine 2021-2023).

首席生物信息学家, 牛津大学(纳菲尔德医学系2018-2021,韦瑟罗尔分子医学研究所2021-2023).

2023+

Professor, Xi’an Jiaotong University.

教授, 西安交通大学.

I also sit on the editorial boards of Frontiers in Genetics (human and medical genomics section; 2020+) and BMC Genomic Data (2022+).