Description

Mach2qtl performs QTL analysis based on imputed dosages/posterior_probabilities. Mach2qtl was developed by Goncalo Abecasis at the University of Michigan. Mach2qtl website.

Mach2qtl is intended to be used interactively on Helix. If large numbers of Mach2qtl jobs are required (> 2-3 simultaneous jobs), or the jobs are expected to run for a long time, the Biowulf cluster may be more appropriate. Please contact the Helix staff (staff@helix.nih.gov) if you have questions about where to run Mach2qtl.

Version

v108

Sample session

helix%  /usr/local/mach2qtl/bin/mach2qtl  -d sample.dat -p sample.ped -i sample.mlinfo --probfile sample.mlprob --dominant --recessive  > test.out

helix% more test.out
Mach2Qtl V1.0.8 (2009-12-16) -- QTL Association Mapping with Imputed Allele Counts
(c) 2007 Goncalo Abecasis, Yun Li


The following parameters are in effect:

Available Options
         Phenotypic Data : --datfile [sample.dat], --pedfile [sample.ped]
   Imputed Allele Counts : --infofile [sample.mlinfo], --dosefile [],
                           --probfile [sample.mlprob]
        Analysis Options : --useCovariates [ON], --quantileNormalization,
                           --dominant [ON], --recessive [ON], --additive [ON]
                  Output : --samplesize

FITTED MODELS (for covariate adjusted residuals)
======================================================
          Trait     Raw Mean Raw Variance         Mean     Variance
        lg10CRP      0.52383      0.13666     -0.00000      0.13180

Loading marker information ...
   1001 markers will be analyzed

Processing prob file ... 

Executing first [ass analysis of imputed genotypes ... 
Executing second pass analysis of imputed genotypes ... 

                        INFORMATION FROM .info FILE             QTL ASSOCIATION ADDITIVE            QTL ASSOCIATION DOMINANT
            QTL ASSOCIATION RECESSIVE        
                        =====================================   =================================   ========================
=========   =================================
TRAIT                   MARKER           ALLELES  FREQ1  RSQR   EFFECT2  STDERR  CHISQ     PVALUE   EFFECT   STDERR  CHISQ  
   PVALUE   EFFECT   STDERR  CHISQ     PVALUE
lg10CRP                 rs9977217        A,G      .3950  .9805   0.020   0.018   1.2320     0.267    0.021   0.026   0.6858 
   0.4076   -0.035   0.035   1.0426    0.3072   
lg10CRP                 rs9977094        A,C      .9894  .3454   0.179   0.133   1.8130    0.1782    0.786   1.022   0.5923 
   0.4415   -0.192   0.143   1.8036    0.1793   

[...etc...]


 Those who expect to use Mach2qtl frequently can add the executable to their path by adding one of the following commands to their ~/.cshrc (csh or tcsh users) or ~/.bashrc (bash users) files.


setenv PATH /usr/local/mach2qtl/bin:$PATH (csh or tcsh)


PATH=/usr/local/mach2qtl/bin:$PATH; export PATH (bash)

Documentation

The README file from this package is available at /usr/local/mach2qtl/README and is displayed below:

mach2qtl
========
QTL analysis based on imputed dosages/posterior_probabilities


Options:
--------

	--datfile : merlin marker info file for phenotypes (required)

	--pedfile : merlin pedigree file for phenotypes (required)

	--infofile : mach1 output .info or .mlinfo file (required)

	--dosefile : mach1 output .dose or .mldose file (required if probfile not available)

	--probfile : mach1 output .prob or .mlprob file (required for non-additive model)

	--useCovariates : adjust for covariates (default is ON)

	--quantileNormalization : apply inverse normal transformation to quantitative trait(s) (default if OFF)

	--dominant : dominant model (default if OFF)

	--recessive : recessive model (default if OFF)

	--additive : additive model (default if ON)

Sample command line & Example
-----------------------------
	
	see sub-folder examples/

Coding
======
For recessive model: AL1/AL1                    1
                     AL1/AL2 & AL2/AL2          0

For dominant model:  AL2/AL2                    1
                     AL1/AL2 & AL1/AL1          0

For additive model:  AL1			0
		     AL2			1
		     (notice the negative sign in "double effect = -numerator / denominator;"

Procedure:
----------
Calculate residuals (if related, use VC)
Perform residual-SNP association

Selected Output
----------------
Raw Mean/Variance vs Mean/Variance
        former for y (could be transformed)
        latter for residuals (even if no residuals, Raw Mean and Mean could be different b/c latter would be centered and thus would be ZERO)


ref:
----
Li Y, Willer CJ, Sanna S, and Abecasis GR (2009). Genotype imputation. Annu Rev Genomics Hum Genet. 10: 387-406. 
Chen WM and Abecasis GR (2007). Family-based association tests for genomewide association scans. Am J Hum Genet 81:913-26