make_drug_target {DrDimont}

Reformat drug-target-interaction data

Description

Function to transform input data to required input format for run_pipeline. Here the data that is needed to define drug-target interactions is formatted. When the reformatted output is passed to run_pipeline as drug_target_interactions argument, the differential integrated drug response score can be calculated for all the supplied drugs in interaction_table.

Usage

make_drug_target(target_molecules, interaction_table, match_on)

Arguments

target_molecules

[string] Name of layer containing the drug targets. This name has to match the corresponding named item in the list of layers supplied to run_pipeline.

interaction_table

[data.frame] Has to contain two columns. A column called 'drug_name' containing names or identifiers of drugs. And a column with a name that matches an identifier in the layer supplied in 'target_molecules'. Additional columns will be ignored in the pipeline. For example, if drugs target proteins and an identifier called 'ncbi_id' was supplied in layer creation of the protein layer (see make_layer), this column should be called 'ncbi_id' and contain the corresponding IDs of protein-drug targets. Any other ID present in the constructed layer could also be used.

match_on

[string] Column name of the data frame supplied in 'interaction_table' that is used for matching drugs and target nodes in the graph (e.g. 'ncbi_id').

Value

Named list of the input parameters in input format of run_pipeline.

Examples

data(drug_gene_interactions)

example_drug_target_interactions <- make_drug_target(target_molecules='protein',
                                        interaction_table=drug_gene_interactions,
                                        match_on='gene_name')