This menu gathers together those quality control, data manipulation, and modeling processes that are particularly useful for copy number analysis. These processes require tall data sets. (See Tall and Wide Data Sets.) Experimental Design Data Set (EDDS) use varies by process. Refer to their individual descriptions for more information.
|
Adjusting copy number intensity or count data by subtracting control sample intensities or counts, either within-subject or across sets
|
|
|
Displaying univariate distribution results for variables, with the option of computing and overlaying density estimates
|
|
|
Standardizing values of numeric variables
|
|
|
Computing correlations between numeric variables, principal components of the correlation matrix, an outlier analysis, and a variance components decomposition
|
|
|
Binning observations into groups, reducing the total number of rows in a data set
|
|
|
Fitting a one-way repeated-measures ANOVA linear model to rows of the input data set, using all combinations of different effects as distinct groups
|
|
|
Fitting a one-way repeated-measures analysis-of-variance linear model to pairs of input data set rows
Important: The bivariate observations are assumed to have the same variance, and a correlation is modeled between them. If the mean or covariance structure is more complex than this, then you should run the Mixed Model Analysis process instead.
Note: This process can be computationally intensive for large data sets, but runs faster than Mixed Model Analysis for identical models.
|
|
|
Finding potential breakpoints of copy number changes across a chromosome using recursive partitioning
|
See the JMP Genomics Starter main page for other process categories.