EP 203: Building the tools behind modern genomics with Jonathan Marchini of Regeneron

EP 203: Building the tools behind modern genomics with Jonathan Marchini of Regeneron

🎙 Jonathan Marchini 👥 942 📅 September 4, 2025 ⏱ 43 min 👁 281 📄 expert opinion 🧭 2026-08-17
Available in: English (current) Français

Keywords

imputationexome sequencinggenotype arraysstatistical geneticscomputational genomics

Summary

In this episode of The Genetics Podcast, host Patrick Short interviews Jonathan Marchini, Head of Statistical Genetics and Machine Learning at the Regeneron Genetics Center. Marchini recounts his career path from teaching mathematics in rural Tanzania to becoming a pioneer in statistical genomics at Oxford, where he contributed to foundational projects like the HapMap, Wellcome Trust Case Control Consortium, and the 1000 Genomes Project. He discusses the development of genotype imputation methods, which have become standard in the field. The conversation then shifts to his work at Regeneron, where he leads efforts to analyze massive exome datasets, including the landmark million-exome paper. Marchini explains the computational challenges of scaling association studies and the development of tools like REGENIE and REMA to handle these data efficiently. He shares insights on why Regeneron prioritizes exome sequencing combined with imputation over whole-genome sequencing for discovery, citing cost-effectiveness and comparable power. The discussion also covers the role of polygenic risk scores in clinical trials and drug development, and where AI adds value in genomics. Marchini emphasizes the importance of user-friendly software and the ongoing need for better phenotyping and rare variant interpretation.

188 words

Critical Evaluation

Value of the Information & Strength of the Argument

The value of the information is high, as it provides an insider’s perspective on the development of key computational methods in genomics and the practical challenges of analyzing large-scale genetic data. Marchini’s arguments are well-reasoned and grounded in his extensive experience, particularly his justification for exome sequencing plus imputation over whole-genome sequencing, which he supports with evidence from his own work. The discussion is balanced, acknowledging limitations and open questions, such as the clinical translation of polygenic risk scores.

88 words

Title / Content Match

The title accurately reflects the content: a detailed conversation about building computational tools for genomics, with a focus on Jonathan Marchini's career and current work at Regeneron.

Quality & Reliability

8/10

The speaker is a leading expert in statistical genetics with a long track record of developing widely used methods. The discussion is grounded in his direct experience and references to published work, but it is primarily an expert opinion interview rather than a systematic review or original study.

Chapters

Cited Sources

  • Regeneron Genetics Center — Mentioned as the center where Jonathan Marchini works and where large-scale exome sequencing is conducted.
  • Million exome paper — Referenced as a landmark publication from the Regeneron Genetics Center, showcasing the scale of their exome sequencing efforts.

Concurring Sources

  • Million exome paper — The paper supports the claims about the scale and findings of exome sequencing at Regeneron.

External References

Contribution & Novelties

The interview provides unique insights into the evolution of computational methods in genomics, from the HapMap era to current large-scale exome analysis. Marchini’s perspective on the trade-offs between exome sequencing and whole-genome sequencing, and the development of tools like REGENIE and REMA, offers valuable knowledge for researchers in the field.

Pour aller plus loin :

  • Genotype imputation — Provides background on the method that Marchini helped pioneer.
  • REGENIE — The software tool developed at Regeneron for scalable association studies.
  • Polygenic risk score — Explains the concept discussed in the context of clinical applications.

93 words

Radar Profile

The radar profile shows high scores across all dimensions, indicating a well-rounded and reliable source of information. The strongest aspects are the quantity and quality of information, reflecting the depth of expertise and the breadth of topics covered.

Reliability 8/10