Pack: Data mining on KUP data

Description:

No description available

Description:

No description available

Description:

Plots a (modified) heatmap of the fibrosis gene data describing different genes used in various studies on fibrosis. The input dataset is prepared in the "prepare-fibrosis-data" RM workflow. Possible values in the input data are: 1 (a gene is Up regulated), -1 (Down regulated), 0 (Unmodified) and -2 (missing value). It requires the "gplots" R package.

Description:

Plots an attribute similarity matrix computed from the provided data frame. The input dataset is prepared in the "prepare-gene-correlation" RM workflow and is assumed to contain genes as rows, and gene functions and locations as columns. In the visualized matrix columns correspond to gene locations and rows correspond to gene functions. The supported types of attribute similarities are: the covariance, the Hamming distance, and a modified Hamming distance that focuses on columns with 1s. It requires "libs.R" to be in the same directory.

Description:

R library used by other R scripts.

Description:

Compressed directory from RM repository containing data used by the different workflows and R scirpts.

Description:

Prepares gene data for fibrosis by joining datasets from different studies on fibrosis across various organs including the kidney.

Description:

Prepares data for gene correlation analysis.

Description:

Generates frequent itemsets and association rules from the gene data.

Frequent itemsets correspond to different gene characteristics (locations and functions) that frequently occur together across different genes. The quality of a frequent itemset is given by its support which is the fraction of genes for which a given combination of gene characteristics hold; the longer the itemset the lower its support.

Based on the frequent itemset this WF also generates association rules which have some itemset in a premise and some in the conclusion. The quality of a rule is given by its confidence.