Consistency of heritability estimation from summary statistics in high-dimensional linear models
David Azriel, Samuel Davenport, Armin Schwartzman · 2025
Read Consistency of heritability estimation from summary statistics in high-dimensional linear models by David Azriel, Samuel Davenport, Armin Schwartzman (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-11144.txt.
Introduction
David Azriel, Samuel Davenport, and Armin Schwartzman examine when the two standard tools for estimating heritability from genome study summary data can be trusted. Heritability here means the share of outcome variation explained by a huge number of genetic markers together. The common tools regress marker statistics on correlation scores between markers. Writing in twenty twenty-five, the team studies the large-sample behavior of both tools and of the tweaks used in practice, such as weighting and standardization, for both bell-shaped and general data.
Two conditions, nicknamed weak dependence and bounded-kurtosis effects, turn out to be enough for the basic estimators to be consistent. For bell-shaped data those conditions are also necessary. Careful truncation keeps weighting harmless, but routine standardization can bias results when the conditions fail. Population stratification biases every version, a warning the authors underline.
Catalog record
| CREATOR | David Azriel, Samuel Davenport, Armin Schwartzman |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 232 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-11144