PLCO Germline Materials for Integrated Whole-Genome Discovery of Lung Cancer Susceptibility with the Sherlock-Lung Study

Principal Investigator

Name
Maria Teresa Landi

Degrees
M.D., Ph.D.

Institution
National Cancer Institute, National Institutes of Health

Position Title
Senior Investigator & Senior Advisor for Genetic Epidemiology

Email
landim@nih.gov

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
2026-0117

Initial CDAS Request Approval
Jul 8, 2026

Title
PLCO Germline Materials for Integrated Whole-Genome Discovery of Lung Cancer Susceptibility with the Sherlock-Lung Study

Summary
Lung cancer remains a leading cause of cancer mortality, and inherited susceptibility is incompletely characterized—particularly for rare variants, structural variants, and non-coding regulatory regions that are not well captured by genotyping arrays or whole-exome sequencing alone. The PLCO Cancer Screening Trial biorepository offers an exceptional, deeply annotated population-based resource with linked demographic and clinical data.
We propose to use PLCO germline materials and integrate them with the Sherlock-Lung cohort (~5,300 whole-genome sequencing [WGS] samples) and additional large-scale resources that we are actively analyzing in collaboration with DCEG investigators (including Diptavo Dutta, Jianxin Shi, and Wendy Wong) and Cancer Genomics Research Laboratory (CGR), to enable comprehensive germline analyses spanning coding and non-coding regions. Specifically, based on the feasibility assessment, we will request 3,177 primary lung cancer cases and 6,354 non-cancer controls (total 9,531 samples) for long-read WGS. In addition, we will request 750 controls for short-read WGS, preferentially all from participants of East Asian ancestry when available; 100 of these controls are already included within the 6,354 long-read controls.
Importantly, we will leverage genotyping data that has already been generated by PLCO for ancestry inference and linkage to existing GWAS resources. Our primary laboratory activity will be performing short- and long-read WGS on the PLCO specimens, and we have obtained special funding that covers the sequencing cost of these samples. WGS will enable discovery of common and rare single-nucleotide variants, indels, structural variants, and non-coding regulatory variants associated with lung cancer susceptibility. We will harmonize processing and analysis with Sherlock-Lung WGS to improve discovery power and facilitate cross-cohort analysis. We will also integrate external resources, including long-read WGS from 23andMe, short-read WGS from the Million Veteran Program (MVP), and meta-analysis with UK Biobank, Genomics England, and All of Us, to bring the total comparative and replication scale to approximately 17,920 case and 27,785 control samples across datasets.
A central objective of this project is to integrate WGS-based discovery with existing GWAS-derived polygenic risk scores (PRS) to achieve a more complete understanding of lung cancer susceptibility, spanning common and rare variants as well as coding and non-coding genomic contributions. In parallel, we have established the infrastructure and operational capacity through CGR as part of the Sherlock-Lung study to receive and store biospecimens, conduct sequencing, and carry out the comprehensive downstream analyses required for this work.

Aims

Aim 1: Generate short- and long-read whole-genome sequencing data from PLCO germline specimens (3,177 lung cancer cases and 6,354 non-cancer controls), together with short-read WGS from 750 matched controls, to discover coding and non-coding susceptibility variants.
- Perform long-read WGS on PLCO germline materials to identify common and rare SNVs/indels, structural variants, and non-coding regulatory variants associated with lung cancer susceptibility, and generate short-read WGS on an additional matched control subset to support ancestry-focused analyses and platform harmonization.
- Apply standardized QC, including cross-platform concordance checks in the 100 overlapping controls, and harmonized processing to ensure robust downstream analyses and cross-cohort integration.

Aim 2: Integrate PLCO WGS with Sherlock-Lung (~5,300 WGS), long-read WGS from 23andMe, short-read WGS from MVP, and meta-analysis datasets from UK Biobank, Genomics England, and All of Us to improve discovery, replication, and generalizability of inherited risk signals.
- Harmonize variant calling, joint genotyping, structural-variant detection, and functional annotation pipelines across cohorts and sequencing modalities, with platform-aware analyses for short-read and long-read data.
- Use combined analyses, replication, and meta-analysis to prioritize high-confidence coding and non-coding susceptibility variants and assess consistency across cohorts and ancestries (where available).

Aim 3: Combine WGS-based findings with existing PLCO genotyping/GWAS resources and GWAS-derived PRS to model lung cancer susceptibility across common/rare and coding/non-coding genetic architectures.
- Leverage existing PLCO genotyping and PRS to quantify how WGS-identified rare and structural variants contribute to risk beyond common-variant polygenic burden.
- Develop integrated analytic frameworks that jointly evaluate PRS and WGS-derived rare variant effects to refine biological insight and improve risk stratification.

Collaborators

Maria Teresa Landi (National Cancer Institute, National Institutes of Health)
Huu Phuc Hoang (National Cancer Institute, National Institutes of Health)
Diptavo Dutta (National Cancer Institute, National Institutes of Health)
Jianxin Shi (National Cancer Institute, National Institutes of Health)
Shuk Wan Wendy Wong (National Cancer Institute, National Institutes of Health)