{
  "id": 18428595,
  "title": "European LD proxies lose tagging for 43% of GWAS leads in the African panel: a chromosome 22 analysis with SwiftLD, an 826-fold faster, byte-identical vcftools LD",
  "url": "https://urgent.news/2026/11/05/european-ld-proxies-lose-tagging-for-43-of-gwas-leads-in-the-african",
  "topic": "science",
  "section": "Science",
  "published": "2026-11-05T00:00:00.000Z",
  "source": {
    "name": "bioRxiv",
    "slug": "biorxiv",
    "url": "https://www.biorxiv.org/content/10.64898/2026.11.04.760946v1?rss=1"
  },
  "original_language": "en",
  "account": "A study conducted on chromosome 22 revealed that European linkage disequilibrium (LD) proxies lagged behind in 43% of genome-wide association studies (GWAS) leads when used in the African panel. Using SwiftLD, a lightning-fast byte-identical alternative to vcftools, researchers demonstrated the speed and accuracy of this new tool.\n\nSwiftLD, a drop-in replacement for vcftools' pairwise LD modes, outperformed its competitors in terms of speed. On three regions of the 1000 Genomes chromosome 22, SwiftLD processed 503 to 2,504 samples with outputs identical to vcftools, all while being 826 times faster on average (geometric mean at two threads on an AVX-512 CPU). The program also outperformed PLINK 2, the fastest alternative, writing LD output more efficiently on all five regions tested.\n\nOne megabase of chromosome 22 with 503 samples took vcftools 29 seconds, whereas SwiftLD completed the task in a mere 0.049 seconds. On an ordinary AVX2 laptop, SwiftLD proved to be 102 times faster than vcftools when using 12 threads. The researchers ensured the program's correctness by using coding agents known as \"Claude Opus 5.5,\" which required all changes to maintain byte-identical outputs to vcftools. An extensive 167,749 differential tests across three machines confirmed that no output byte differed.\n\nIn the context of GWAS leads, the best European LD proxy kept the r2 value at 0.8 for 57.4% of leads in the African panel and maintained between 85.6% to 92.7% in other panels (98.1% in a European split-half control). The researchers have made SwiftLD's code available on GitHub, along with a two-command reproduction guide for interested researchers.",
  "summary": "SwiftLD is a drop-in replacement for the pairwise linkage disequilibrium (LD) modes of vcftools 0.1.17, --hap-r2 and --geno-r2, which population-genetic pipelines use to compute r2, D and D' between nearby variants. On three regions of 1000 Genomes chromosome 22 with 503 to 2,504 samples it wrote the same output as vcftools, byte for byte, and was 826-fold faster (geometric mean at two threads on…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}