PLINK

PLINK
Name	PLINK
Developer	Shaun Purcell, Christopher Chang, and others
Released	2005
Latest release	1.90/2.0 series
Programming language	C/C++
Operating system	Linux, macOS, Windows
License	Open-source

PLINK is an open-source whole-genome association analysis toolset widely used in human genetics, population genetics, and genomics. It provides utilities for quality control, association testing, population stratification, linkage disequilibrium, and data conversion across genotype and phenotype formats. The software has influenced large-scale projects and consortia by enabling reproducible pipelines for variant-level analyses and cohort-level quality checks.

Overview

PLINK is a command-line toolkit developed for single nucleotide polymorphism (SNP) data analysis and supports genotype and phenotype processing for cohort studies. It integrates with workflows common to projects such as the 1000 Genomes Project, UK Biobank, HapMap, ENCODE Project, and clinical cohorts from institutions like the Broad Institute and the Wellcome Sanger Institute. Its functionality is often combined with software and resources including VCFtools, bcftools, GATK, EIGENSOFT, and REGENIE to build end-to-end pipelines used by groups such as the International HapMap Consortium and the Genome Aggregation Database teams.

History and Development

Initial development began in the early 2000s by researchers seeking scalable tools for genome-wide association studies (GWAS), motivated by landmark studies like the early GWAS on age-related macular degeneration and collaborations between groups at the Massachusetts General Hospital and Harvard Medical School. Contributions and major updates involved researchers affiliated with institutions such as Stanford University, University of Cambridge, and the UCLA Fielding School of Public Health. Over time, versions evolved from the original PLINK 1.0 to the widely used 1.07/1.90 line and the PLINK 2.0 rewrite aimed at performance improvements. The project has been discussed at conferences including the American Society of Human Genetics annual meeting and cited in publications from journals like Nature Genetics and The American Journal of Human Genetics.

Features and Functionality

PLINK implements a broad suite of features for genotype data, including SNP and sample quality control, basic and logistic regression association tests, allele frequency computation, Hardy–Weinberg equilibrium testing, identity-by-descent and relatedness estimation, principal component analysis for population structure, linkage disequilibrium metrics, and haplotype-based tests. These functions complement methods developed in software such as KING, ADMIXTURE, STRUCTURE, and fastSTRUCTURE. PLINK also supports permutation testing and meta-analysis preparatory steps relevant to consortia like the GIANT Consortium and the Psychiatric Genomics Consortium.

File Formats and Data Handling

PLINK native formats such as PED/MAP and BED/BIM/FAM support compact storage and rapid I/O for genotype matrices, and conversion utilities enable interoperability with formats like VCF and dosage formats used by imputation tools such as IMPUTE2 and Minimac. Data handling routines include sampling, merging, strand alignment, allele recoding, and phenotype covariate management—common preprocessing steps before analyses conducted by projects like the Alzheimer's Disease Sequencing Project and the Cancer Genome Atlas. Many biobank pipelines use PLINK to convert between platform-specific outputs (e.g., arrays from Illumina or Affymetrix) and centralized analysis formats.

Common Analyses and Workflows

Typical workflows begin with sample- and variant-level quality control, removal of population outliers using PCA relative to reference panels such as 1000 Genomes Project or HapMap, kinship/relatedness filtering, association testing (linear or logistic regression), and post hoc evaluation of inflation metrics like genomic control lambda and QQ plots—often in combination with visualization tools or R packages from groups at Bioconductor and the R Project for Statistical Computing. PLINK is used in GWAS pipelines for traits analyzed by teams at institutions including Johns Hopkins University, Massachusetts Institute of Technology, and the National Institutes of Health.

Performance and Scalability

Performance improvements in successive releases addressed memory footprint, multithreading, and on-disk binary formats to enable analyses of large cohorts such as UK Biobank and national biobanks like those from Estonia or Iceland-based projects. PLINK 2.0 introduced optimizations for dense genotype matrices and multi-allelic handling to scale to hundreds of thousands of samples and millions of variants, complementing high-performance computing workflows on systems provided by organizations like Amazon Web Services, the European Bioinformatics Institute, and university HPC clusters.

Adoption and Impact on Genomics

PLINK has become a staple tool referenced across thousands of publications and used by consortia including the Psychiatric Genomics Consortium, GIANT Consortium, and the International HapMap Consortium. Its role in standardizing QC and association workflows has influenced data sharing practices and reproducible research in genomics, affecting translational studies at centers like Mayo Clinic, Cleveland Clinic, and pharmaceutical genome programs at companies such as Genentech and Pfizer. Training workshops at conferences hosted by groups like the Wellcome Trust and the European Society of Human Genetics frequently include PLINK in curriculum for genetic epidemiology and computational genomics.

Category:Genetics software