%\VignetteIndexEntry{iCARE Vignette}
%\VignettePackage{iCARE}
%\VigetteDepends{iCARE}

\documentclass[a4paper]{article}
\begin{document}

\title{iCARE(Individualized Coherent Absolute Risk Estimators) Package}
\maketitle

<<start>>=
library(iCARE)
@

\section*{Example 1.A}

Load the breast cancer data.
<<load-data>>=
data("bc_data", package="iCARE")
@

In this example, we will estimate the risk of breast cancer in ages 50-80.
A SNP-only model is fit, with no specific genotypes supplied for estimation.
The population disease rates are from SEER.
<<model-1a>>=
res_snps_miss = computeAbsoluteRisk(model.snp.info = bc_15_snps, 
                      model.disease.incidence.rates = bc_inc,
                      model.competing.incidence.rates = mort_inc, 
                      apply.age.start = 50, 
                      apply.age.interval.length = 30,
                      return.refs.risk=TRUE)
@

Compute a summary of the risks and visualize the results
<<summary-1a, fig=TRUE>>=
summary(res_snps_miss$risk)
summary(res_snps_miss$refs.risk)
plot(density(res_snps_miss$risk), lwd=2, 
  main="SNP-only Risk Stratification: Ages 50-80", 
  xlab="Absolute Risk of Breast Cancer")
@

\section*{Example 1.B}

In this example, we will again estimate the risk of breast cancer in ages 50-80.
This time however, three specific genotypes are supplied for estimation (with some missing data).
The argument return.refs.risk = TRUE, includes the referent dataset risks be included in results.
<<model 1b>>=
res_snps_dat = computeAbsoluteRisk(model.snp.info = bc_15_snps, 
                                model.disease.incidence.rates = bc_inc,
                                model.competing.incidence.rates = mort_inc, 
                                apply.age.start = 50, 
                                apply.age.interval.length = 30, 
                                apply.snp.profile = new_snp_prof, 
                                return.refs.risk = TRUE)
names(res_snps_dat)
@

Visualize the Results
<<plot-1b, fig=TRUE>>=
plot(density(res_snps_dat$refs.risk), lwd=2, 
   main="Referent SNP-only Risk Distribution: Ages 50-80",
   xlab="Absolute Risk of Breast Cancer")
abline(v=res_snps_dat$risk, col="red")
legend("topright", legend="New Profiles", col="red", lwd=1)
@

\section*{Example 2}

In this example, we will estimate the risk of breast cancer in ages 50-80 by
fitting a model with two risk factors and 15 SNPs, with
three specific covariate profiles supplied for estimation (with some missing data).
First, the model.cov.info argument is created. 

<<ex-2>>=
v1 <- list(name="famhist", type="continuous")
v2 <- list(name="parity", type="factor", levels=0:4, ref=0)
bc_model_cov_info <- list(v1, v2)
@

Now fit the model
<<fit-2>>=
res_covs_snps = computeAbsoluteRisk(model.formula=caco~famhist+as.factor(parity), 
                                         model.cov.info=bc_model_cov_info,
                                         model.snp.info=bc_15_snps,
                                         model.log.RR=bc_model_log_or,
                                         model.ref.dataset=ref_cov_dat,
                                         model.disease.incidence.rates=bc_inc,
                                         model.competing.incidence.rates=mort_inc, 
                                         model.bin.fh.name="famhist",
                                         apply.age.start=50, 
                                         apply.age.interval.length=30,
                                         apply.cov.profile=new_cov_prof,
                                         apply.snp.profile=new_snp_prof, 
                                         return.refs.risk=TRUE)
@

Display details of the fit
<<fit-2>>=
print(res_covs_snps$details)
@

\section*{Session Information}
<<sessionInfo>>=
sessionInfo()
@ 

\end{document}