SQL & PL/SQL :: Customized Query Builder Architecture

I got a task to send some alters based on the user query data .As user is not a SQL expert, so we have to provide him a query builder. Now to store user query, I decided to build a data model and keeping in mind the reusability of previously made queries (so that we can use them directly in further queries).Like a user made two queries:

1. Find patient id, name, sugar_level, pulse_rate, weight for those patients whose diabitic level is greater than 'X'
2. Find patient id, pulse_rate for all those patients whose diabitic level is greater than 'X' and weight > 90

So first query can be directly used in second query. (Just a very simple example)But I am pretty much confused with the decision of columns for each table to store query and how to relate them.

Tables I am able to identify is :

1. Table to store query level details, i.e. query_id, who created, query_short_name, query_description, operator used (SET, GROUP BY, HAVING, etc), distinct indicator, etc.
2. select clause table or columns selected table
3. From clause table
4. table to store group by clause
5. table to store having clause

I am trying to store seprate parts of the query into seprate tables, but not getting which columns to have in each table, how to relate them... and after that, for a given query_id, procedure to construct a whole query, and then validate/parse and execute it using dbms_sql and scheduler.