I'm trying to essentially get all of the data under a specific high level node and return it as a nested JSON. I've been messing around with it for a while, and I could probably make something work with a lot of OPTIONAL MATCHES and WITH, but it looks gigantic, and it's not clean.

In addition, I run into issues with duplicates. In the schema below, if x1 has multiple x1 nodes, and multiple y1 child nodes, then I get duplicated x1 nodes in my result set. Which is not ideal.

Should I just run separate queries, and combine in code? Should I try to go down this path of writing a massive query and try to combine in a single JSON on the fly?

I'll define my schema first, so I have something like

(Movie)-[rel1]-(Actor)-[x]

where [x] can be:
[x1]-(node1)-[y1]
    [y1]-(something)

[x2]-(node2)
    [y1]-(something)

[x3]-(node2)
    [y1]-(something)

[x4]-(node2)
[x5]-(node2)

and what I want is something like:

{

Movie: 
   ...movie properties,
   Actor: [
       x1: [
           {
               ...x1 properties,
               y1: [
                   {...y1 properties},
                   {...y1 properties}
               ]
           },
           { another x1 node with properties, and its children of y1's}
       ]
       x2... (like above)
       x3... (like above)
   ]

}

This is the query I've been playing around with. I've tried a few different iterations.

OPTIONAL MATCH (Movie)-[rel1]->(Actor)-[x1]->(node1)-[y1]-(something1)
WITH Movie, Actor, { node1: COLLECT(node1), something1: COLLECT(something1)} as node1

// continue writing same query as above, with x2 and y2, etc

RETURN {
  Movie: Movie,
  Actor: COLLECT({
    Actor: Actor,
    node1: node1, // this returns duplicates
  })
}