Scaling genetic discovery through automated post-GWAS interpretation

Abstract

Genome-wide association studies (GWAS) have made association discovery systematic and cumulative, yet evidence for interpreting associated loci remains selective, heterogeneous and difficult to accumulate across studies. Here we present Post-GWAS Intelligence (PGI), which couples context-dependent post-GWAS analysis to a common representation of analysis-derived evidence. Its multi-agent engine, VariantAgent, executes workflows, records tool-derived results in standardized variant-centred units and generates trait-level interpretations linked to their analytical basis and evidential status. Complementary benchmarks showed that VariantAgent outperformed tested general-purpose and biomedical agents, with its largest advantage when success required both a correct answer and a valid analytical trace. Across 34 published GWAS, VariantAgent achieved a mean per-study recovery of 82.1% for directly comparable findings. The largest gains lay downstream of association mapping, where uniform analysis extended published interpretations and identified regulatory relationships at established risk loci not previously reported in the source studies. Analysis of 1,041 GWAS yielded 530,949 evidence units and corresponding reports, which PGI organized into a cross-trait catalogue. Catalogue interrogation resolved an apparent cross-trait overlap into distinct signals and trait-specific biological hypotheses. This work demonstrates how PGI can extend genetic discovery beyond association mapping by enabling biological evidence to accumulate across studies.

Sum-h2, enabling genetic discovery for deep learning-derived phenotypes through a fast evaluation framework and arena of performance
T. Xia, Z. Xie, X. Zhao +5
bioRxiv24 Sep 2026PreprintCC BY-NCFull textBioinformatics & computational biologybioRxiv
An automated genetic screen identifies modulators of stress-induced sleep in Caenorhabditis elegans
Z. J. Li, H. Honarpisheh, S. Kutagulla +4
bioRxiv16 May 2026PreprintAll rights reservedFull textGenetics & genomicsbioRxiv
Synthesis-aware generative design in trillion-scale chemical spaces for automated drug discovery
S. Seo, Y. Sung, S.-Y. Hwang +12
bioRxiv1 Oct 2026PreprintCC BYFull textBioinformatics & computational biologybioRxiv
Optimizing Cognitive Phenotyping for Genetic Discovery in Alzheimer's Disease: A Simulation-Based Evaluation of Mini-Mental State Examination-Based Decline Measures
Ruyu Shi, Eleanor Feingold, M. Ilyas Kamboh +1
bioRxiv16 Oct 2026PreprintAll rights reservedFull textBioinformatics & computational biologybioRxivPMC
Discovery of a Virulent Population of Meloidogyne graminicola Overcoming the Rice Resistance Gene MG1 and Population Genetic Analysis Based on mtCOI Sequences
Ao Xu, Borong Lin, Qiuling Huang +4
Plants (Basel, Switzerland)1 Oct 2026Peer-reviewedCC BYFull textPlant biologyEurope PMC

Cite

BibTeX RIS

BibTeX

@misc{qiao2026scaling,
  title = {{Scaling genetic discovery through automated post-GWAS interpretation}},
  author = {L. Qiao and S. Gu and Y. Yang and Y. Han and X. Li and X. Wan and X. Zhang and Q. Zhang and T. Hu and Y. Zhang and F. Chen and J. He and J. Wang and Z. Tang and P. Ye and L. Wang and E. Long and X. Wang and L. Bai and F. Yu},
  howpublished = {bioRxiv},
  year = {2026},
  month = oct,
  doi = {10.64898/2026.10.23.761278},
  url = {https://www.biorxiv.org/content/10.64898/2026.10.23.761278v1},
  note = {CC BY},
}

RIS

TY  - UNPB
TI  - Scaling genetic discovery through automated post-GWAS interpretation
AU  - L. Qiao
AU  - S. Gu
AU  - Y. Yang
AU  - Y. Han
AU  - X. Li
AU  - X. Wan
AU  - X. Zhang
AU  - Q. Zhang
AU  - T. Hu
AU  - Y. Zhang
AU  - F. Chen
AU  - J. He
AU  - J. Wang
AU  - Z. Tang
AU  - P. Ye
AU  - L. Wang
AU  - E. Long
AU  - X. Wang
AU  - L. Bai
AU  - F. Yu
PY  - 2026
DA  - 2026/10/23
PB  - bioRxiv
DO  - 10.64898/2026.10.23.761278
UR  - https://www.biorxiv.org/content/10.64898/2026.10.23.761278v1
AB  - Genome-wide association studies (GWAS) have made association discovery systematic and cumulative, yet evidence for interpreting associated loci remains selective, heterogeneous and difficult to accumulate across studies. Here we present Post-GWAS Intelligence (PGI), which couples context-dependent post-GWAS analysis to a common representation of analysis-derived evidence. Its multi-agent engine, VariantAgent, executes workflows, records tool-derived results in standardized variant-centred units and generates trait-level interpretations linked to their analytical basis and evidential status. Complementary benchmarks showed that VariantAgent outperformed tested general-purpose and biomedical agents, with its largest advantage when success required both a correct answer and a valid analytical trace. Across 34 published GWAS, VariantAgent achieved a mean per-study recovery of 82.1% for directly comparable findings. The largest gains lay downstream of association mapping, where uniform analysis extended published interpretations and identified regulatory relationships at established risk loci not previously reported in the source studies. Analysis of 1,041 GWAS yielded 530,949 evidence units and corresponding reports, which PGI organized into a cross-trait catalogue. Catalogue interrogation resolved an apparent cross-trait overlap into distinct signals and trait-specific biological hypotheses. This work demonstrates how PGI can extend genetic discovery beyond association mapping by enabling biological evidence to accumulate across studies.
ER  -

Abstract

Related

Cite