aop_backdoor: Backdoor Causal Network Analysis for AOPs
In aop: Adverse Outcome Pathway Analysis

Description Usage Arguments Details Value Examples

Performs a backdoor causal network analysis to idenitfy nodes/key events which are sufficient to infer causality.

1	aop_backdoor(aop_graph, ke_coord, ao_coord, measureable_nodes = NULL)

`aop_graph`	a `graphNEL` object that encodes the AOP. Typically, this would be output from the `convert_aop_to_graph function`.
`ke_coord`	typically this is the molecular initiating event node, but really, this is any node that you want as the starting/source point. For instance, this is normally the point at which exposure to a stressor is going to enter the AOP.
`ao_coord`	typically this is the adverse outcome node.
`measureable_nodes`	this param is not used yet. In the future this node will be a `vector` of the nodes where an assay is available to measure the node. In a future release this param will focus the backdoor algorithm on finding only those nodes for which measurements can actually be taken, as opposed to causal nodes regardless of our ability to measure them. This allows for the assumption that AOP key events may or may not be measureable.

This function performs Pearl's backdoor analysis. Whereas Pearl was interested in identifying nodes which need to be measured to make a causal statement, we are interested in identifying those nodes/key events which need to measured to say that an adverse outcome is likely to occur. It's essentially the same thing as Pearl, only a slightly different interpretation.

causal_nodes vector a vector of the names of the causal nodes.

steatosis_json_file <- system.file("extdata", "steatosis_aop_json.cyjs",
package = "aop")
steatosis_aop <- convert_cytoscape_to_aop(steatosis_json_file)
steatosis_aop_graph <- convert_aop_to_graph(steatosis_aop)
aop_backdoor(steatosis_aop_graph, "391", "388")