Description Usage Arguments Value Examples
Main Function for Meta analysis: microarray & RNAseq
The MetaDE
is a function to identify genes associated with the
response/phenoype of interest (can be either group, continuous or survival)
by integrating multiple studies(datasets).
The main input consists of raw expression data.
1 2 3 4 |
data |
is a list of K elements, where K is the number of studies, each element is a microarray or RNAseq expression matrix with G rows and N columns, where G is number of matched genes and N is the sample size. |
clin.data |
is a list of K elements, each element includes is a clinical data frame with N rows and p columns, where N is the sample size and p is the number of clinical variables (main response included). |
data.type |
is a character indicating the data type of the elements
in |
resp.type |
is a character indicating the response type of the
|
response |
is one column name of |
covariate |
are the clinical covariates you wish to adjust for in the DE analysis, can be a vector of column names or NULL. |
ind.method |
is a character vector to specify the method used to test if there is association between the gene expression and outcome variable. For "twoclass" or "multiclass" response, must be one of "limma", "sam" for "continuous" data type and "edgeR", "DESeq2" or "limmaVoom" for "discrete" data type. For "continuous" response, use "pearsonr" or "spearmanr". For "survival", use "logrank". |
meta.method |
is a character to specify the Meta-analysis method used to combine the p-values, effect sizes or ranks. Available methods include: "maxP","maxP.OC","minP","minP.OC","Fisher","Fisher.OC","AW", roP","roP.OC", "Stouffer","Stouffer.OC","SR","PR","minMCC","FEM","REM","rankProd". |
paired |
is a logical vecter of size K to indicate whether the study is paired design? |
rth |
is the option for roP and roP.OC method. rth means the rth smallest p-value. |
REM.type |
is the option for "REM" method only, choose from "HS","HO", "DL", "SJ", "EB" or "RML". |
asymptotic |
is a logical value indicating whether asymptotic distribution should be used. If FALSE, permutation will be performed. |
tail |
is a character string specifying the alternative hypothesis, must be one of "abs" (default), "low" or "high". |
parametric |
is a logical values indicating whether the parametric methods is chosen to calculate the p-values in meta-analysis. |
nperm |
is the number of permutations. Applicable when |
select.group: |
for two-class comparison only, specify the two groups for comparison when the group factor has more than two levels. |
ref.level: |
for two-class/multi-class comparison only, specify the reference level of the group factor. |
seed: |
Optional initial seed for random number generator. |
a list with components:
stat: a matrix with rows representing genes. It is the statistic for the selected meta analysis method of combining p-values.
pval: the p-value from meta analysis for each gene for the above stat.
FDR: the FDR of the p-value for each gene for the above stat.
AW.weight: The optimal weight assigned to each dataset/study for
each gene if the 'AW
' method was chosen.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 | data('Leukemia')
data('LeukemiaLabel')
data <- Leukemia
K <- length(data)
clin.data <- lapply(label, function(x) {data.frame(x)} )
for (k in 1:length(clin.data)){
colnames(clin.data[[k]]) <- "label"
}
select.group <- c('inv(16)','t(15;17)')
ref.level <- "inv(16)"
data.type <- "continuous"
ind.method <- c('limma','limma','sam')
resp.type <- "twoclass"
paired <- rep(FALSE,length(data))
meta.method <- "Fisher"
meta.res <- MetaDE(data=data,clin.data = clin.data,
data.type=data.type,resp.type = resp.type,
response='label',
ind.method=ind.method, meta.method=meta.method,
select.group = select.group, ref.level=ref.level,
paired=paired,tail='abs',parametric=TRUE)
meta.method <- "Fisher.OC"
meta.res <- MetaDE(data=data,clin.data = clin.data,
data.type=data.type,resp.type = resp.type,
response='label',
ind.method=ind.method, meta.method=meta.method,
select.group = select.group, ref.level=ref.level,
paired=paired,tail='high',parametric=FALSE,nperm=100)
meta.method <- "FEM"
meta.res <- MetaDE(data=data,clin.data = clin.data,
data.type=data.type,resp.type = resp.type,
response='label',
ind.method=ind.method, meta.method=meta.method,
select.group = select.group, ref.level=ref.level,
paired=paired, tail='abs')
meta.method <- "REM"
REM.type <- "HO"
meta.res <- MetaDE(data=data,clin.data = clin.data,
data.type=data.type,resp.type = resp.type,
response='label',
ind.method=ind.method, meta.method=meta.method,
select.group = select.group, ref.level=ref.level,
paired=paired,
REM.type=REM.type,tail='abs')
meta.method <- "SR"
meta.res <- MetaDE(data=data,clin.data = clin.data,
data.type=data.type,resp.type = resp.type,
response='label',
ind.method=ind.method, meta.method=meta.method,
select.group = select.group, ref.level=ref.level,
paired=paired,tail='abs',parametric=FALSE,nperm=100)
meta.method <- 'minMCC'
meta.res <- MetaDE(data=data,clin.data = clin.data,
data.type=data.type,resp.type = resp.type,
response='label',
ind.method=ind.method, meta.method=meta.method,
select.group = select.group, ref.level=ref.level,
paired=paired,tail='abs',parametric=FALSE,nperm=100)
meta.method <- "AW"
meta.res <- MetaDE(data=data,clin.data = clin.data,
data.type=data.type,resp.type = resp.type,
response='label',covariate = NULL,
ind.method=ind.method, meta.method=meta.method,
select.group = select.group, ref.level=ref.level,
paired=paired, rth=NULL,
REM.type=NULL,tail='abs',parametric=TRUE)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.