This version contains a whole new suite of tools for handling copy number variation data, for both rare (segmental) and common variants. As well as a number of minor fixes and improvements, the web manual is now available as a single PDF file.

Version 1.01 released 27 January 2008

This contains a number of additions and fixes, largely minor in nature, relative to 1.00. In particular, there is updated proxy association/imputation command syntax, output, features and default settings. Problems introduced in the last release, in QFAM, sliding window haplotype analysis, and the --hap-impute function are fixed. Also, some convenience functions have been added, e.g. to update a map file (--update-map) and create dummy variables for categorical covariates (--dummy-coding). In addition, various parts of the online documentation have been updated.

Version 1.00 released 5 December 2007

After a relatively long period since 0.99s was released, version 1.00, the first "stable" release, provides several important fixes and updates as well as several completely new features. Although several new features are designated as being in a beta stage of development, releasing as version 1.00 (as opposed to 0.99t) signifies both a reasonable level of maturity for many parts of this package as well as the fact that PLINK will, for the foreseeable future, still be subject to regular change (in the form of additions and updates as well as bug fixes).

The main new features are

Conditional haplotype-based association tests: a set of routines have been added that implement and extend most features offered in the no-longer-supported WHAP software package.
A simple facility to clump groups of SNP-based results across one or more files based on linkage-disequilibrium
Beta development version of a full SNP imputation routine (i.e. to impute and test ungenotyped SNPs given a reference such as the Phase 2 HapMap)

Version 0.99s released 27 July 2007

There are several entirely new sets of features in version v0.99s. This release was scheduled to be the v1.00 release, but given the number of new features added, it seemed appropriate to have one more beta release, to iron out any major problems before the main release.

The main additions are:

A set of proxy association methods, that are designed to present single SNP associations in their haplotypic context: for example, is a single SNP association also seen in the surrounding haplotypes? Also, the basic single SNP test is reframed in terms of a haplotype test, which can have some advantages with respect to non-random genotyping failure.
There is now a web-based lookup function, to quickly give report a set of different types of annotation data for particular SNPs.
PLINK now has some degree of extensibility, via a R plugin feature. Users can define their own test statistic using the R language, which can then be easily embedded within a PLINK run.
Data can now be loaded in "long format" (in which one row corresponds to a genotype, rather than a person or a SNP).
Changed the implementation of the --homozyg-* functions, that looks for extended stretches of homozygosity.
Several other minor additions and bug-fixes. Most notable bug fixes are for the DFAM test, having covariates and genotypic tests in the --linear and --logistic functions; a problem with the --bmerge option.

There have also been a number of improvements and fixes made to the gPLINK GUI (now version v1.00). In addition, several improvements and additions relevant to viewing PLINK output files have been made to the Haploview program (version 4.0, release candidate 2) available from here.

Version 0.99r released 29 April 2007

Main additions include the DFAM and QFAM tests, in addition to several bug fixes and data management options.

Manuscript describing PLINK in press

A manuscript is in press (American Journal of Human Genetics). Please do not distribute this pre-publication version.

This document last modified Wednesday, 11-Jun-2008 18:14:43 EDT