MATCH

This describes a pattern that the database should match. This is probably the most important piece of the query as it is a structural component that always starts your queries.

MATCH (me:Person)-[:KNOWS]->(friend)

WHERE

This filters results that are found in the match for specific criteria.

WHERE me.name = "My Name" AND me.age > 18

RETURN

This returns results. You can either return paths, nodes, relationships, or their properties—or an aggregate of the mentioned parameters. This is another structural component, as all read queries and most write queries will return some data.

RETURN me.name, collect(friend), count(*) as friends

WITH

This passes results from one query part to the next. Much like RETURN, but instead of including data in the result set, it will be passed to the following part of the query. It transforms and aggregates results and also separates READ and UPDATE statements.

ORDER BY

SKIP

LIMIT:

This sorts and paginates the results.

ORDER BY friends DESC SKIP 10 LIMIT 10

CREATE

This creates nodes and relationships with their properties.

CREATE (p:Person), (p)-[:KNOWS {since: 2010}]-> (me:Person {name:"My Name"})

CREATE UNIQUE

This fixes graph structures by only creating structures if they do not yet exist.

MERGE

This matches or creates semantics by using indexes and locks. You can specify different operations in case of a MATCH (part of the pattern already existed) or on CREATE (pattern did not exist yet).

MERGE (me:Person {name:"My Name"})

ON MATCH me SET me.accessed = timestamp()

ON CREATE me SET me.age = 42

SET, REMOVE

This updates properties and labels on nodes and/or relationships.

SET me.age = 42SET me:Employee REMOVE me.first_name REMOVE me:Contractor

DELETE

It deletes nodes and relationships.

MATCH (me)

OPTIONAL MATCH (me)-[r]-() DELETE me, r