modelOutcome            package:MergeMaid            R Documentation

_C_o_m_p_a_r_e _r_e_g_r_e_s_s_i_o_n _c_o_e_f_f_i_c_i_e_n_t_s _a_c_r_o_s_s _s_t_u_d_i_e_s

_D_e_s_c_r_i_p_t_i_o_n:

     Given a set of merged studies, this function calculates study
     specific regression coefficients for each gene.

_U_s_a_g_e:

        modelOutcome(x,outcome,outcome2=NULL,method=c("linear","logistic","cox"),...)

_A_r_g_u_m_e_n_t_s:

       x: Object of class mergeExpressionSet.

  method: Method specifies the model used to generate coefficients.  At
          this time only linear regression, logistic regression, and
          Cox hazard rates are implemented.

outcome, outcome2: The format for the outcome variable depends on the
          model used.  For linear regression, outcome should be a
          continous response variable, for logistic regression, it
          should be a binary response variable, and for Cox hazard
          rates it should be time of event.  Outcome 2 is currently
          used only in the calculation of hazard rates, and should be a
          binary variable indicating censoring status for each subject.
          If outcome is a vector of length equal to number of studies,
          then each element represents the column in the ExpressionSet
          phenoData slot for that study.  If outcome is a list, then
          each list element should have actual outcome data for the
          corresponding study. 

     ...: Not implemented at this time

_V_a_l_u_e:

     The output is a mergeCoeff object.

_S_e_e _A_l_s_o:

     'modelOutcome', 'mergeCoeff-class'

_E_x_a_m_p_l_e_s:

       if(require(Biobase) & require(MASS) & require(survival)){
       data(mergeData)
       merged  <- mergeExprs(sample1,sample2,sample3)

       log.coeff  <- modelOutcome(merged,outcome=c(1,1,1),method="logistic")
       plot(coeff(log.coeff))
       
       linear.coeff  <- modelOutcome(merged[1:2],outcome=c(3,3),method="linear")
       plot(zscore(linear.coeff),xlab="study 1",ylab="study 2")

       event1<-rbinom(100,1,.5)
       event2<-rbinom(50,1,.5) 
       event3<-rbinom(70,1,.5)

       out1<-rnorm(100,5,1)
       out2<-rnorm(50,5,1)
       out3<-rnorm(70,5,1)

       out<-list(out1,out2,out3)
       even<-list(event1,event2,event3)

       cox.coeff<-modelOutcome(merged,outcome2=even,outcome=out,method="cox")
       plot(coeff(cox.coeff))

       }

