A data harmonization pipeline to leverage external controls and boost power in GWAS

Chen, Danfeng; Tashman, Katherine; Palmer, Duncan S.; Neale, Benjamin M.; Roeder, Kathryn; Bloemendal, Alex; Churchhouse, Claire; Ke, Zheng Tracy

doi:10.1093/hmg/ddab261

Cited by 7 publications

(13 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Controls should be matched to cases with the same genotyping platform, the same variant calling, quality-control metrics and analysis pipeline, and imputed together with the same variants and reference panel [ 43 ]. Close attention should be paid to possible structural differences in the data caused by different laboratories and/or study populations [ 17 – 20 ]. These issues are generic to borrowing controls, irrespective of whether one borrows 1 or 100 control(s)/case.…”

Section: Discussionmentioning

confidence: 99%

Increase in power by obtaining 10 or more controls per case when type-1 error is small in large-scale association studies

Katki

Berndt

Machiela

et al. 2023

BMC Med Res Methodol

View full text Add to dashboard Cite

Background The rule of thumb that there is little gain in statistical power by obtaining more than 4 controls per case, is based on type-1 error α = 0.05. However, association studies that evaluate thousands or millions of associations use smaller α and may have access to plentiful controls. We investigate power gains, and reductions in p-values, when increasing well beyond 4 controls per case, for small α. Methods We calculate the power, the median expected p-value, and the minimum detectable odds-ratio (OR), as a function of the number of controls/case, as α decreases. Results As α decreases, at each ratio of controls per case, the increase in power is larger than for α = 0.05. For α between 10–6 and 10–9 (typical for thousands or millions of associations), increasing from 4 controls per case to 10–50 controls per case increases power. For example, a study with power = 0.2 (α = 5 × 10–8) with 1 control/case has power = 0.65 with 4 controls/case, but with 10 controls/case has power = 0.78, and with 50 controls/case has power = 0.84. For situations where obtaining more than 4 controls per case provides small increases in power beyond 0.9 (at small α), the expected p-value can decrease by orders-of-magnitude below α. Increasing from 1 to 4 controls/case reduces the minimum detectable OR toward the null by 20.9%, and from 4 to 50 controls/case reduces by an additional 9.7%, a result which applies regardless of α and hence also applies to “regular” α = 0.05 epidemiology. Conclusions At small α, versus 4 controls/case, recruiting 10 or more controls/cases can increase power, reduce the expected p-value by 1–2 orders of magnitude, and meaningfully reduce the minimum detectable OR. These benefits of increasing the controls/case ratio increase as the number of cases increases, although the amount of benefit depends on exposure frequencies and true OR. Provided that controls are comparable to cases, our findings suggest greater sharing of comparable controls in large-scale association studies.

show abstract

Section: Discussionmentioning

confidence: 99%

Increase in power by obtaining 10 or more controls per case when type-1 error is small in large-scale association studies

Katki

Berndt

Machiela

et al. 2023

BMC Med Res Methodol

View full text Add to dashboard Cite

show abstract

“…For R 2 bins in between, AF diff progressively increases as imputation quality decreases (Supp Fig 1 ). For example, the 514 EUR AFFY samples with WGS had 2.10% of SNPs (7,064,852) with AF diff ≥ 1% in S1, decreasing to 1.98% in S2. Looking at a more relaxed threshold of AF diff ≥ 2%, S1 had 0.39% of SNPs meeting this criterion whereas S2 slightly increased to 0.56% (Supp Table 2).…”

Section: Tsim Is Robust Against Imputation-derived Errorsmentioning

confidence: 99%

“…When combining cohorts, batch effects can arise due to differences in genotyping platform. To reduce these effects, the current practice for combining cohorts prior to imputation is to use genotyped single nucleotide polymorphisms (SNPs) that are shared between cohorts 1,4–7 . This approach not only reduces the number of SNPs available for analysis but may also adversely affect genotype imputation by decreasing accuracy.…”

Section: Introductionmentioning

confidence: 99%

“…However, this method requires the external cohort to have WGS available which may not be feasible for many underrepresented populations, particularly in regions with limited funding and resources. Another study sought to harmonize multiple control cohorts by removing low-quality genotyped SNPs, then merging genotypes after a single imputation 7 . These SNPs were identified through an iterative process of GWAS using genotyping platform as the outcome, to identify spurious genome-wide significant loci, and re-imputation on array genotypes, to identify low-quality SNPs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Accurate cross-platform GWAS analysis via two-stage imputation

Greenberg,

Reynolds,

McNulty

et al. 2024

Preprint

View full text Add to dashboard Cite

In genome-wide association studies (GWAS), combining independent case-control cohorts has been successful in increasing power for meta and joint analyses. This success sparked interest in extending this strategy to GWAS of rare and common diseases using existing cases and external controls. However, heterogeneous genotyping data can cause spurious results. To harmonize data, we propose a new method, two-stage imputation (TSIM), where cohorts are imputed separately, merged on intersecting high-quality variants, and imputed again. We show that TSIM minimizes cohort-specific bias while controlling imputation-derived errors. Merging arthritis cases and UK Biobank controls using TSIM, we replicated known associations without introducing false positives. Furthermore, GWAS using TSIM performed comparably to the meta-analysis of nephrotic syndrome cohorts genotyped on five different platforms, demonstrating TSIM's ability to harmonize heterogeneous genotyping data. With the plethora of publicly available genotypes, TSIM provides a GWAS framework that harmonizes heterogeneous data, enabling analysis of small and case-only cohorts.

show abstract

“…al. have noted similar problems where an “aggregation of controls from multiple sources is challenging due to batch effects, difficulty in identifying genotyping errors and the use of different genotyping platforms” [ 192 ].…”

Section: Proposed Solutionsmentioning

confidence: 99%

The evolution of Big Data in neuroscience and neurology

et al. 2023

View full text Add to dashboard Cite

Neurological diseases are on the rise worldwide, leading to increased healthcare costs and diminished quality of life in patients. In recent years, Big Data has started to transform the fields of Neuroscience and Neurology. Scientists and clinicians are collaborating in global alliances, combining diverse datasets on a massive scale, and solving complex computational problems that demand the utilization of increasingly powerful computational resources. This Big Data revolution is opening new avenues for developing innovative treatments for neurological diseases. Our paper surveys Big Data’s impact on neurological patient care, as exemplified through work done in a comprehensive selection of areas, including Connectomics, Alzheimer’s Disease, Stroke, Depression, Parkinson’s Disease, Pain, and Addiction (e.g., Opioid Use Disorder). We present an overview of research and the methodologies utilizing Big Data in each area, as well as their current limitations and technical challenges. Despite the potential benefits, the full potential of Big Data in these fields currently remains unrealized. We close with recommendations for future research aimed at optimizing the use of Big Data in Neuroscience and Neurology for improved patient outcomes.

show abstract

A data harmonization pipeline to leverage external controls and boost power in GWAS

Cited by 7 publications

References 36 publications

Increase in power by obtaining 10 or more controls per case when type-1 error is small in large-scale association studies

Increase in power by obtaining 10 or more controls per case when type-1 error is small in large-scale association studies

Accurate cross-platform GWAS analysis via two-stage imputation

The evolution of Big Data in neuroscience and neurology

Contact Info

Product

Resources

About