Title



     mrmodal -- Modal estimator for summary data



Syntax



        mrmodal varname_gd varname_gdse varname_gp varname_gpse [if] [in] [, options]



    options               Description
    ----------------------------------------------------------------------------------------------
      level(#)            set confidence level; default is level(95)
      nome                NOME assumption
      nosave              Do not save density and vector of IV estimates in Mata
      phi(#)              value of phi (for bandwidth)
      reps(#)             number of bootstrap replications to obtain standard error
      seed(#)             seed for random number generator for bootstrapping to obtain standard
                            error
      weighted            weighted IV estimates



Description



    mrmodal implements the zero modal estimator of Hartwig et al. (2017) for use with summary
    level data (i.e. reported genotype-disease [SNP-outcome] and genotype-phenotype [SNP-exposure]
    association estimates and their standard errors for individual genotypes).



    Standard errors are obtained by parametric bootstrapping.



    varname_gd is a variable containing the genotype-disease (SNP-outcome) association estimates.



    varname_gdse is a variable containing the genotype-disease (SNP-outcome) association estimate
    standard errors.



    varname_gp is a variable containing the genotype-phenotype (SNP-exposure) association
    estimates.



    varname_gpse is a variable containing the genotype-phenotype (SNP-exposure) association
    estimate standard errors.



Options



    level(#); see [R] estimation options.



    nome specifies the NOME (no measurement error in the genotype-phenotype [SNP-exposure]
        associations) assumption.



    nosave specifies that the density of the IV estimates and column vector of IV estimates should
        not be saved in Mata. If not specified these are saved in Mata as mrmodal_densityiv and
        mrmodal_g respectively.



    phi(#) specifies the parameter phi which is used in the calculation of the bandwidth for the
        density estimation. Default is phi = 1, other values commonly chosen are 0.25 and 0.5.



    reps(#) specifies the number of bootstrap replications for obtaining the standard error. The
        default is 1000 replications.



    seed(#) specifies the initial value of the random-number seed.  The default is the current
        random-number seed. Specifying seed(#) is the same as typing set seed # before issuing the
        command; see set_seed.



    weighted weight the instrumental variable estimates.



Examples



    Using the data provided by Do et al. (2013).



    Setup
        . use https://raw.github.com/remlapmot/mrrobust/master/dodata, clear



    Select observations (p-value with exposure < 10^-8)
        . gen byte sel1 = (ldlcp2 < 1e-8)



    Investigate what is a good value of phi to use (we want a smooth density plot)
        . mrmodalplot chdbeta chdse ldlcbeta ldlcse if sel1==1



    Simple mode estimator
        . mrmodal chdbeta chdse ldlcbeta ldlcse if sel1==1



    Simple mode estimator with reproducible standard error
        . mrmodal chdbeta chdse ldlcbeta ldlcse if sel1==1, seed(12345)



    Weighted mode estimator
        . mrmodal chdbeta chdse ldlcbeta ldlcse if sel1==1, weighted



    Simple mode estimator with NOME assumption
        . mrmodal chdbeta chdse ldlcbeta ldlcse if sel1==1, nome



    Weighted mode estimator with NOME assumption
        . mrmodal chdbeta chdse ldlcbeta ldlcse if sel1==1, weighted nome



Stored results



    mrmodal stores the following in e():



    Scalars        
      e(k)                number of instruments
      e(phi)              value of phi
      e(reps)             number of (bootstrap) replications



    Macros         
      e(cmd)              mrmodal
      e(cmdline)          command as typed



    Matrices       
      e(b)                coefficient vector
      e(V)                variance-covariance matrix of the estimates



References



    Do et al. Common variants associated with plasma triglycerides and risk for coronary artery
        disease. Nature Genetics, 2013, 45, 1345-1352.  DOI



    Hartwig FP, Davey Smith G, Bowden J.  Robust inference in two-sample Mendelian randomisation
        via the zero modal pleiotropy assumption. International Journal of Epidemiology, 2017, 46,
        6, 1985-1998.  DOI



Author



INCLUDE help mrrobust-author