INTERVAL BioResource

Summary

INTERVAL is a UK blood-donor research resource originally established through a randomized trial of blood-donation intervals. Its linked genetic, complete-blood-count and multi-omic measurements support studies of blood-cell biology, molecular-trait prediction and statistical-method development [1].

Role in Latent-Factor Analysis

The flashfmZero protocol uses European-ancestry INTERVAL participants with classical and non-classical complete-blood-count traits measured on Sysmex XN analyzers. The example trait dataset contains 43,059 participants, with maximum trait missingness of 32%, and provides observed-trait GWAS summary statistics, a phenotypic covariance matrix and regional association and LD data for the SMIM1 locus.[1]

Classical measures cover hemoglobin, major blood-cell concentrations and average red-cell or platelet volumes. Non-classical traits include cell-type-specific granularity and nucleic-acid-content measurements, creating the high-dimensional correlation structure used in multivariate latent-factor genetic analysis.[1]

Reuse Across the Wiki

INTERVAL also supports genetic prediction models for transcriptomic, proteomic and metabolomic traits and has been used to develop or validate circulating biomarker resources. Cohort-specific sample selection, assay platforms and ancestry composition should be checked for each downstream analysis.

The primary flashfmZero study used 18,310 participants with complete measurements for individual-level latent-factor GWAS, up to 43,059 participants for summary-statistics blood analyses, and up to 40,849 participants for 184 Nightingale NMR traits. The close agreement of factor contributions between the 18k and 43k analyses supported use of pairwise-complete covariance estimates.[3]

Citations

[1] Astle, Butterworth and Asimit (2025), flashfmZero protocol [2] Moore et al. (2014), INTERVAL trial protocol [3] Zhou et al. (2025), blood-cell and metabolic latent-factor analyses