findAttractor: Finding attractor using the seed gene
In weiyi-bitw/cafr: Attractor Metagenes Finding Algorithm

Description Usage Arguments Details Value Note Author(s) References See Also Examples

Find a converged attractor based on the seed gene provided.

1
2

findAttractor(data, vec, a=5, maxIter=100, epsilon=1E-14, bin=6, so=3, rankBased=FALSE, negateMI = TRUE, verbose=TRUE)
findCorrAttractor(data, vec, a=10, maxIter=100, epsilon=1E-14, rankBased=FALSE)

`data`	An expression matrix with genes in the rows, samples in the columns.
`vec`	A vector of expression values of the seed gene.
`a`	Exponent of the mutual information, used to create weight vector for metagenes.
`maxIter`	Max number of iterations.
`epsilon`	Threshold of convergence.
`bin`	Number of bins used when estimate mutual information (default=6).
`so`	Spline order used when estimate mutual information (default=3).
`rankBased`	When `TRUE`, convert the expression values into ranks.
`negateMI`	When `TRUE`, negate the mutual information if the two vectors have negative momentum.
`verbose`	When `TRUE`, it will show the top 20 genes of the metagene in each iteration.

Given a seed gene, findAttractor finds the attractor by iteratively creating metagenes using the weighted average of all genes in data. The weight of each gene is determined by (MI)^a where MI is the mutual information between the gene and the metagene. The metagene in the initial iteration is the seed itself. The iterative process continues until convergence is reached. In each iteration, the program calculates the sum of squares SS of the differences between the vector of MIs in the current iteration and that of MIs in the previous iteration:

SS = (I_1 - I'_1)^2 + (I_2 - I'_2)^2 + …

If SS < epsilon, the process is deemed converged. If the process cannot reach convergence within maxIter, NULL is returned. If negateMI is TRUE, MI between the genes with negative momentum with metagene will be negated, and the weight of this gene is set to 0.

findCorrAttractor does the same thing as findAttractor, but using correlation coefficients instead of MI as association measure.

Returns a vector of MIs between all the genes in the dataset and the converged metagene. If the process does not converge within the maximum number of iteration assigned, it will return NULL.

Missing values are not allowed as the input to the function in the current version.

Wei-Yi Cheng

Wei-Yi Cheng, Tai-Hsien Ou Yang and Dimitris Anastassiou, Biomolecular events in cancer revealed by attractor metagenes, PLoS Computational Biology, Vol. 9, Issue 2, February 2013.

getCorr, getMI,probeSummarization

# Load the toy dataset extracted from TCGA OV data
data(ov)

# find attractor using CENPA as a seed
x <- ov["CENPA",]
a <- findAttractor(ov, x)

# display the top 10 genes in the attractor
sort(a, decreasing=TRUE)[1:10]