Artificial Components Detection of Differentially Expressed Genes

Description

This package provides a multivariate inferential analysis method for detecting differentially expressed genes in gene expression data. It uses artificial components, close to the data's principal components but with an exact interpretation in terms of differential genetic expression, to identify differentially expressed genes while controlling the false discovery rate (FDR). The methods on this package are described in the article Multivariate Method for Inferential Identification of Differentially Expressed Genes in Gene Expression Experiments by Acosta (2015).

Details

Index:

ac              Artificial Components for Gene 
                Expression Data
acde-package    Artificial Components Detection of
                Differentially Expressed Genes
bcaFDR          BCa Confidence Upper Bound for the FDR.
fdr             False Discovery Rate Computation
phytophthora    Gene Expression Data for Tomato Plants
                Inoculated with _Phytophthora infestans_
plot.STP        Plot Method for Single Time Point Analysis
plot.TC         Plot Method for Time Course Analysis
print.STP       Print Method for Single Time Point Analysis
print.TC        Print Method for Time Course Analysis
qval            Q-Values Computation
stp             Single Time Point Analysis for Detecting
                Differentially Expressed Genes
tc              Time Course Analysis for Detecting
                Differentially Expressed Genes

Author(s)

Juan Pablo Acosta, Liliana Lopez-Kleine

Maintainer: Juan Pablo Acosta <jpacostar@unal.edu.co>

References

Acosta, J. P. (2015) Strategy for Multivariate Identification of Differentially Expressed Genes in Microarray Data. Unpublished MS thesis. Universidad Nacional de Colombia, Bogot'a.

Examples

## Single time point analysis for 500 genes with 10 treatment 
## replicates and 10 control replicates
n <- 500; p <- 20; p1 <- 10
des <- c(rep(1, p1), rep(2, (p-p1)))
mu <- as.matrix(rexp(n, rate=1))
Z <- t(apply(mu, 1, function(mui) rnorm(p, mean=mui, sd=1)))
### 5 up regulated genes
Z[1:5,1:p1] <- Z[1:5,1:p1] + 5
### 10 down regulated genes
Z[6:15,(p1+1):p] <- Z[6:15,(p1+1):p] + 4

resSTP <- stp(Z, des)
resSTP
plot(resSTP)

## Time course analysis for 500 genes with 10 treatment 
## replicates and 10 control replicates
tPts <- c("h0", "12h", "24h")
n <- 500; p <- 20; p1 <- 10
Z <- vector("list", 3)
des <- vector("list", 3)
for(tp in 1:3){ des[[tp]] <- c(rep(1, p1), rep(2, (p-p1))) }
mu <- as.matrix(rexp(n, rate=1))
### h0 time point (no diff. expr.)
Z[[1]] <- t(apply(mu, 1, function(mui) rnorm(p, mean=mui, sd=1)))
### h12 time point (diff. expr. begins)
Z[[2]] <- t(apply(mu, 1, function(mui) rnorm(p, mean=mui, sd=1)))
#### Up regulated genes
Z[[2]][1:5,1:p1] <- Z[[2]][1:5,1:p1] + 
    matrix(runif(5*p1, 1, 3), nrow=5)
#### Down regulated genes
Z[[2]][6:15,(p1+1):p] <- Z[[2]][6:15,(p1+1):p] + 
    matrix(runif(10*(p-p1), 1, 2), nrow=10)
### h24 time point (maximum differential expression)
Z[[3]] <- t(apply(mu, 1, function(mui) rnorm(p, mean=mui, sd=1)))
#### 5 up regulated genes
Z[[3]][1:5,1:p1] <- Z[[3]][1:5,1:p1] + 5
#### 10 down regulated genes
Z[[3]][6:15,(p1+1):p] <- Z[[3]][6:15,(p1+1):p] + 4

resTC <- tc(Z, des)
resTC
summary(resTC)
plot(resTC)

Results

R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> library(acde)
Loading required package: boot
> png(filename="/home/ddbj/snapshot/RGM3/R_BC/result/acde/acde-package.Rd_%03d_medium.png", width=480, height=480)
> ### Name: acde-package
> ### Title: Artificial Components Detection of Differentially Expressed
> ###   Genes
> ### Aliases: acde-package acde
> ### Keywords: package
> 
> ### ** Examples
> 
> ## Single time point analysis for 500 genes with 10 treatment 
> ## replicates and 10 control replicates
> n <- 500; p <- 20; p1 <- 10
> des <- c(rep(1, p1), rep(2, (p-p1)))
> mu <- as.matrix(rexp(n, rate=1))
> Z <- t(apply(mu, 1, function(mui) rnorm(p, mean=mui, sd=1)))
> ### 5 up regulated genes
> Z[1:5,1:p1] <- Z[1:5,1:p1] + 5
> ### 10 down regulated genes
> Z[6:15,(p1+1):p] <- Z[6:15,(p1+1):p] + 4
> 
> resSTP <- stp(Z, des)
> resSTP

Single time point analysis for detecting differentially
expressed genes in microarray data.

Achieved FDR: 0.1%.

Inertia ratio: %.
tstar: 4.726, pi0: 1, B: 100.
 
Differentially expressed genes:
down-reg.  no-diff.   up-reg. 
       10       485         5 

Results: 
     psi1   psi2 Q-value Diff. expr.
1   5.381  7.371   0.000     up-reg.
2   4.866  7.905   0.000     up-reg.
3  10.191  7.609   0.000     up-reg.
4   7.805  6.548   0.000     up-reg.
5   6.220  7.602   0.000     up-reg.
6   9.052 -7.267   0.000   down-reg.
7   5.451 -6.080   0.000   down-reg.
8   5.549 -6.069   0.000   down-reg.
9  11.105 -5.235   0.001   down-reg.
10  2.733 -5.384   0.001   down-reg.
11  3.132 -5.327   0.001   down-reg.
12  4.353 -5.312   0.001   down-reg.
13  6.245 -4.861   0.001   down-reg.
14  2.918 -4.848   0.001   down-reg.
15  2.676 -4.726   0.001   down-reg.
16 -4.354 -1.976   0.248    no-diff.
17 -0.036 -1.755   0.401    no-diff.
18 -0.970 -1.735   0.410    no-diff.
19 -3.052 -1.677   0.430    no-diff.
20 -1.809  1.677   0.430    no-diff.
21 -0.092  1.595   0.442    no-diff.
22  5.513  1.606   0.442    no-diff.
23 -0.571  1.615   0.442    no-diff.
24 -1.236  1.588   0.442    no-diff.
25 -3.005  1.558   0.457    no-diff.
...

*More results are available in the objects:
$ac, $qvalues and $dgenes.
> plot(resSTP)
> 
> ## Time course analysis for 500 genes with 10 treatment 
> ## replicates and 10 control replicates
> tPts <- c("h0", "12h", "24h")
> n <- 500; p <- 20; p1 <- 10
> Z <- vector("list", 3)
> des <- vector("list", 3)
> for(tp in 1:3){ des[[tp]] <- c(rep(1, p1), rep(2, (p-p1))) }
> mu <- as.matrix(rexp(n, rate=1))
> ### h0 time point (no diff. expr.)
> Z[[1]] <- t(apply(mu, 1, function(mui) rnorm(p, mean=mui, sd=1)))
> ### h12 time point (diff. expr. begins)
> Z[[2]] <- t(apply(mu, 1, function(mui) rnorm(p, mean=mui, sd=1)))
> #### Up regulated genes
> Z[[2]][1:5,1:p1] <- Z[[2]][1:5,1:p1] + 
+     matrix(runif(5*p1, 1, 3), nrow=5)
> #### Down regulated genes
> Z[[2]][6:15,(p1+1):p] <- Z[[2]][6:15,(p1+1):p] + 
+     matrix(runif(10*(p-p1), 1, 2), nrow=10)
> ### h24 time point (maximum differential expression)
> Z[[3]] <- t(apply(mu, 1, function(mui) rnorm(p, mean=mui, sd=1)))
> #### 5 up regulated genes
> Z[[3]][1:5,1:p1] <- Z[[3]][1:5,1:p1] + 5
> #### 10 down regulated genes
> Z[[3]][6:15,(p1+1):p] <- Z[[3]][6:15,(p1+1):p] + 4
> 
> resTC <- tc(Z, des)
> resTC

Time course analysis for detecting differentially expressed
genes in microarray data.
 
Inertia ratios (%): 
   t1   t2    t3
 4.59 5.73 14.34

Active timepoint: t3 


TIME POINT: t1.

Achieved FDR: 41.1%.

Inertia ratio: %.
tstar: 1.93, pi0: 1, B: 100.
 
Differentially expressed genes:
down-reg.  no-diff.   up-reg. 
        6       491         3 

Results: 
     psi1   psi2 Q-value Diff. expr.
1  -2.562  2.089   0.411     up-reg.
2  -2.429  2.234   0.411     up-reg.
3  -1.032  2.142   0.411     up-reg.
4  -1.385 -1.938   0.411   down-reg.
5   1.340 -2.020   0.411   down-reg.
6  -2.811 -1.930   0.411   down-reg.
7   6.296 -2.156   0.411   down-reg.
8  -0.763 -2.205   0.411   down-reg.
9  -3.107 -1.949   0.411   down-reg.
10 -0.424 -1.875   0.447    no-diff.
...

*More results are available in the objects:
$ac, $qvalues and $dgenes.


TIME POINT: t2.

Achieved FDR: 1.9%.

Inertia ratio: %.
tstar: 2.472, pi0: 1, B: 100.
 
Differentially expressed genes:
down-reg.  no-diff.   up-reg. 
        4       492         4 

Results: 
     psi1   psi2 Q-value Diff. expr.
1   5.469  4.013   0.000     up-reg.
2   2.107  2.956   0.003     up-reg.
3   0.599  2.739   0.008     up-reg.
4   1.819  2.472   0.019     up-reg.
5   0.860 -2.965   0.003   down-reg.
6   9.735 -2.744   0.008   down-reg.
7  -1.565 -2.571   0.014   down-reg.
8   0.236 -2.569   0.014   down-reg.
9  -2.643 -2.247   0.062    no-diff.
10 11.815 -2.058   0.124    no-diff.
11  0.797  1.997   0.145    no-diff.
12  3.567 -1.960   0.152    no-diff.
13  9.001 -1.909   0.187    no-diff.
14 -0.011  1.872   0.205    no-diff.
15  7.136 -1.808   0.212    no-diff.
16  1.371 -1.817   0.212    no-diff.
17  2.832  1.830   0.212    no-diff.
18 -2.194 -1.825   0.212    no-diff.
...

*More results are available in the objects:
$ac, $qvalues and $dgenes.


TIME POINT: t3.

Achieved FDR: 0%.

Inertia ratio: %.
tstar: 4.967, pi0: 1, B: 100.
 
Differentially expressed genes:
down-reg.  no-diff.   up-reg. 
       10       485         5 

Results: 
     psi1   psi2 Q-value Diff. expr.
1   5.890  6.757   0.000     up-reg.
2   5.180  7.463   0.000     up-reg.
3   6.278  7.055   0.000     up-reg.
4   8.436  6.907   0.000     up-reg.
5   7.250  7.336   0.000     up-reg.
6  10.283 -6.267   0.000   down-reg.
7   3.634 -6.077   0.000   down-reg.
8  11.000 -5.543   0.000   down-reg.
9   3.896 -6.072   0.000   down-reg.
10  2.789 -5.001   0.000   down-reg.
11  2.395 -5.251   0.000   down-reg.
12  7.659 -4.967   0.000   down-reg.
13  5.521 -5.126   0.000   down-reg.
14  6.878 -5.810   0.000   down-reg.
15 12.923 -5.385   0.000   down-reg.
16  0.005  2.244   0.164    no-diff.
17 -0.239  2.183   0.175    no-diff.
18 -4.004 -2.095   0.199    no-diff.
19  2.620  1.915   0.283    no-diff.
20 -0.884 -1.885   0.285    no-diff.
21  0.757 -1.835   0.312    no-diff.
22 -3.109  1.637   0.497    no-diff.
23  1.871  1.519   0.623    no-diff.
24 -1.354  1.520   0.623    no-diff.
25  0.064  1.477   0.678    no-diff.
...

*More results are available in the objects:
$ac, $qvalues and $dgenes.
> summary(resTC)

Time course analysis for detecting differentially
expressed genes in microarray data.
 
Inertia ratios (%): 
   t1   t2    t3
 4.59 5.73 14.34


Active vs complementary time points analysis:

Active timepoint: t3 

Achieved FDR: 0 %.

Differentially expressed genes:
down-reg.  no-diff.   up-reg. 
       10       485         5 


Groups conformation through time analysis:

Differentially expressed genes:
 t2 vs t3           
            down-reg. no-diff. up-reg. Sum
  down-reg.         4        0       0   4
  no-diff.          6      485       1 492
  up-reg.           0        0       4   4
  Sum              10      485       5 500
> plot(resTC)
> 
> 
> 
> 
> 
> dev.off()
null device 
          1 
>