skoyama427

./plink --bfile GWAStutorial --out penn --make-bed
./plink --bfile penn --out qc_penn --maf 0.05 --hwe 0.0000001 --make-bed

awk -F "," 'NR>1 {print $1" 1 "$3" "$4}' GWAStutorial_clinical.csv> PENN.cov
awk -F "," 'NR>1 {print $1" 1 "$2+1}' GWAStutorial_clinical.csv> CAD.pheno
awk -F "," 'NR>1 && $5!~/NA/ {print $1" 1 "$5}' GWAStutorial_clinical.csv> TG.pheno
awk -F "," 'NR>1 && $6!~/NA/ {print $1" 1 "$6}' GWAStutorial_clinical.csv> HDL.pheno
awk -F "," 'NR>1 && $7!~/NA/ {print $1" 1 "$7}' GWAStutorial_clinical.csv> LDL.pheno

./plink --bfile qc_penn --out qc_CAD --pheno CAD.pheno --assoc  --ci 0.95

skoyama427

library(rstan)
library(MASS)

stanmodel <- "

  data {

    int N;
    real x[N];
    vector[2] y[N];

skoyama427

stanmodel <- "

  data {

    int N;
    matrix[N,2] X;

  }

  parameters {

skoyama427

library(mass)
library(mvtnorm)

set.seed(31415)

MU    <- c(10, 30)
SIGMA <- c(20, 40)
RHO   <- -0.7

CVM <- rbind(c(SIGMA[1]^2, SIGMA[1]*SIGMA[2]*RHO),

skoyama427

stancodeLASSO="
data {
  int N;
  int M;
  vector[N] Y;
  matrix[N, M] X;
}
parameters {
  real<lower=0> S;
  vector[M] B;

skoyama427

library(psych)
library(glmnet)

set.seed(1)

x1 <- rnorm(30); x2 <- rnorm(30); x3 <- rnorm(30); e  <- rnorm(30)

X  <- matrix(c(x1, x2, x3), byrow = F, ncol = 3) #explanatory variables
y  <- 3 + 0.01*x1 + 0.01*x2 + 5*x3 + rnorm(30)   #y is a linear combination of x1x2x3
fit <-glmnet(x = X, y = y, alpha = 1)            #fit use glmnet (alpha=1 : LASSO)

skoyama427

[[Session handling]]

<check session>
$ tmux ls

<attach>
$ tmux a -t 2

<detach>
PREFIX d

skoyama427

m <- c(1:9, 3, 3)

structure(sapply(m, sqrt), dim=m))
structure(vapply(m, sqrt, numeric(1)), dim=m))

m[] <- sapply(m, sqrt)
m[] <- vapply(m, sqrt, numeric(1))

skoyama427

set nohlsearch
set cursorline
set mouse=a
set noswapfile
set nobackup
set noundofile
set showmatch
set matchtime=1
set ttimeoutlen=10
set number