Statistical fine-mapping part 1

Statistical fine-mapping part 1

by International Statistical Genetics Workshop

Formal & Physical Sciences Biology PSALife sciences: general issuesPSAXComputational biology
🎙 International Statistical Genetics Workshop 👥 3K 📅 May 27, 2026 ⏱ 13 min 👁 239 📄 lecture 🧭 2026-08-16
Available in: English (current) Français

Keywords

fine-mappingGWASPIPcredible setsLD

Summary

This lecture introduces statistical fine-mapping, a method to identify likely causal variants in genomic regions associated with traits. The speaker contrasts fine-mapping with GWAS, emphasizing that GWAS identifies associations but not causality. Fine-mapping outputs two key quantities: Posterior Inclusion Probability (PIP) and credible sets. PIP quantifies the probability that a SNP is causal given the data, while credible sets capture uncertainty due to linkage disequilibrium (LD). Using real data from the 2022 PGC Schizophrenia study, the lecture illustrates fine-mapping results in three regions: a simple region with two high-PIP SNPs in one credible set, a more complex region with three credible sets, and a highly complex region with very high LD leading to large credible sets and low PIPs. The lecture then discusses factors influencing fine-mapping efficacy and resolution: LD and sample size (citing Schaid et al., 2018), SNP density, missing causal variants, effect size, and number of causal SNPs. The speaker concludes by noting that larger sample sizes and lower LD improve fine-mapping, and that missing causal variants can lead to false positives.

174 words

Critical Evaluation

Value of the Information & Strength of the Argument

The lecture provides valuable insights into the practical application of fine-mapping, using real examples to illustrate concepts. The argumentation is solid, supported by references to a key paper (Schaid et al., 2018) and real data from the PGC study. The speaker clearly explains the role of LD and sample size, and the impact of other factors such as SNP density and missing variants. The presentation is logical and builds on previous lectures, making it accessible to a range of audiences.

Scientific Rigor, Source Quality, Title Accuracy

The lecture demonstrates scientific rigor by referencing a peer-reviewed publication (Schaid et al., 2018) and using real data from a major consortium (PGC). The title accurately reflects the content, which is the first part of a lecture on statistical fine-mapping. The speaker does not overstate claims and acknowledges limitations, such as the difficulty in fine-mapping regions with high LD. Overall, the sources are credible and the content is well-aligned with the title.

167 words

Title / Content Match

The title accurately reflects the content, which is the first part of a lecture on statistical fine-mapping.

Quality & Reliability

8/10

The lecture is presented by an expert in statistical genetics, with clear explanations and references to a peer-reviewed paper (Schaid et al., 2018). The content is well-structured and based on established methods.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

  • Schaid et al., 2018, Nature Reviews Genetics — Simulation study on the effects of LD and sample size on fine-mapping.
  • PGC 2022 Schizophrenia study — Real data used to illustrate fine-mapping results.

Concurring Sources

  • Schaid et al., 2018, Nature Reviews Genetics — The lecture's discussion of LD and sample size aligns with this paper's findings.

Contribution & Novelties

The lecture provides a clear and accessible introduction to statistical fine-mapping, emphasizing practical considerations and real-world examples. It effectively explains the concepts of PIP and credible sets, and highlights factors that influence fine-mapping performance. The use of real data from the PGC study adds credibility and helps viewers understand the challenges in practice.

Pour aller plus loin :

93 words

Radar Profile

The radar profile shows high scores in quantity and quality of information, with a moderate technical level. The lecture is well-balanced, providing both theoretical and practical insights, making it suitable for a broad audience interested in statistical genetics.

Reliability 8/10