XWAS 3.0

XWAS (chromosome X-Wide Analysis toolSet) is a software suite for analysis of the X chromosome in genome-wide association studies and similar types of studies.

The X chromosome plays an important role in complex human traits and diseases, especially those with sexually dimorphic characteristics. Special attention needs to be given to the analysis of X due to its unique inheritance pattern and X-inactivation. These analytical complications have resulted in exclusion or mishandling of the X chromosome in the majority of genome-wide association studies (GWAS) to date. With XWAS, we hope to provide the tools and incentive to accurately incorporate the X chromosome into GWAS, hence enabling discoveries of novel loci implicated in many diseases and in their sexual dimorphism.

The XWAS package includes standard GWAS procedures adapted specifically to X:

  • Optional sex-aware genotype calling for the X chromosome from raw intensity data
  • Quality control tailored to the X chromosome
  • Optional sex-aware imputation for the X chromosome
  • An array of statistical tests (with various options) for association of X-linked markers
  • Association tests for genes on the X chromosome
  • Gene-gene interaction tests for genes on the X chromosome and the autosomes
  • A visualization suite for association results

See below for a full explanation of the XWAS software.

Sign our downloads journal and tell us a little bit more about yourself:





By checking the box you can sign up to receive emails upon the release of new versions, updates, and the implementation of new features (low volume list).


What’s new?

Click here to view XWAS version 3.0 release notes (May 1, 2018).

Selected Recent Related Publications (How to Cite)

1. Please cite these papers if you use the basic QC or association methods:

Gao F, Chang D, Biddanda A, Ma L, Guo Y, Zhou Z, Keinan A. "XWAS: a toolset for genetic data analysis and association studies of the X chromosome." Journal of Heredity. 2015;106(5):6.

Chang D, Gao F, Slavney A, Ma L, Waldman Y, Sams A, Billing-Ross P, Madar A, Spritz R, Keinan A. "Accounting for eXentricities: Analysis of the X Chromosome in GWAS Reveals X-Linked Genes Implicated in Autoimmune Diseases.PLoS ONE. 2014;9(12):e113684.

2. Please cite this paper if you use the X-inactivation-based methods:

Ma L, Hoffman G, Keinan A. "X-inactivation informs variance-based testing for X-linked association of a quantitative trait." BMC Genomics. 2015;16:241.

3. Please cite this paper if you use gene-based tests for interaction:

Ma L, Clark AG, Keinan A. "Gene-based testing of interactions in association studies of quantitative traits."PLoS Genet.. 2013;9:e1003321

About the Software

XWAS 3.0 begins with optional sex-aware genotype calling for the X chromosome from raw intensity data based on an Affymetrix genotype array. The procedure detects and reports genotypes and their likelihoods for each SNP-by-individual combination. XWAS can also summarize the differences from another set of genotype calls (potentially produced by a genotype caller that was not sex-aware), and visualize the genotyping intensity clusters, which can illustrate the importance of accounting for sex in calling genotypes in plates of mixed-sex. 

The software then provides extensive quality control procedures that are tailored to X (e.g. filter variants with artifactual sex difference in allele frequency or level of missingness in control; apply QC separately to males and females and consider only the intersection of remaining SNPs; remove pseudoautosomal regions). It removes related individuals in the sample with a fine-grained resolution to preserve as many individuals as possible. Followed is optional sex-aware imputation based on IMPUTE2, which scales the effective size of the population (Ne) by 25% for X.

The software offers an array of statistical tests of association for X-linked markers while correcting for population stratification based on EIGENSTRAT and controlling for any other covariates. Tests include:

  1. Standard tests of association with disease risk, with male hemizygotes considered equivalent to female homozygotes by counting each male allele twice (0/2 coding), including the variant of this test implemented by Clayton, or by considering each male allele as a single copy (0/1 coding)
  2. Testing for association in each sex separately before combining into a single statistic using one of two methods: Fisher’s method that accommodates the possibility of sex-differential effect size/direction and is not affected by allele coding (0/2 or 0/1), and Stouffer’s method, which weighs evidence from each sex proportionally to sample size,
  3. Testing for sex-differences in effect size/direction based on the above sex-stratified tests
  4. Motivated by X-inactivation, a test for higher variance of a trait in heterozygous females as compared to homozygous females, as well as a weighted regression test of the mean that accounts for this potential inflation in phenotypic variance due to X-inactivation, and a test that combines the two tests
  5. A modified epistasis test that uses a t-statistic for quantitative traits
  6. Calculation and output of confidence interval of effect size or odds ratio for a subset of the above association tests
  7. X-wide genomic control for a subset of the above association tests
  8. A meta-analysis test which performs both fixed- and random-effects meta-analyses of association results in females or males separately
  9. A convenient command "--run-all" to run several X-related tests at once
  10. For all tests, a gene-based version of the test that combines evidence from all variants in each X-linked genes
  11. Tests for gene-gene interaction, which combine SNP-by-SNP interaction tests between all pairs of SNPs in two or more genes to produce a gene-level test for interaction

The XWAS software package is implemented in C++ and scripts. It includes functions from PLINK. It is freely available for academic usage and offered open source. It is optimized for LINUX, but can be compiled for Windows and MAC OS.

To download our previous releases:


The Keinan Lab continues to develop, support, and release updates of the XWAS package on a regular basis. Ongoing development include gene-based gene-gene interaction testing for case-control studies, parallelized version of IBD for control for related individuals, IBD mapping designed specifically for characteristics of the X chromosome. We also started developing all functionality needed for sequence-based associations studies, starting with optimizing variant and genotype calling for X and X-linked tests for rare variant association studies (RVAS) that build on the unique population genetics of X, which enable more powerful rare variant tests than for the autosomes.

To receive updates about future versions or any bugs, please sign up for our mailing list above. Please also report bugs or ask questions by emailing the XWAS support email (keinanlab.xwas@gmail.com).

Our lab is also involved in collaborative projects, including putting to use our expertise, software, and a variety of case-specific scripts for analysis of the X chromosome and of gene-gene interactions in existing data from association studies. We welcome additional collaborations. Please email Alon Keinan (alon.keinan@cornell.edu) to explore such opportunities.