Over-indexing m2m table for FKs?

Hi all - I would appreciate help on what indices are needed exactly for
a many to many table. This is on 9.0 today, will be 9.2 soon. Here's a
contrived example:

table sidea (a_id)
table sideb (b_id)
table a_to_b (a_id references sidea on delete cascade, b_id references
sideb on delete cascade)

a_to_b is unique, so: create unique index idx ON a_to_b (a_id, b_id)

I have two primary use cases:

1. select sideb.* from a_to_b, sideb where a_to_b.b_id = sideb.b_id and
a_to_b.a_id = X

2. select sidea.*, sideb.* from sidea, a_to_b, sideb where sidea.a_id =
a_to_b.a_id and a_to_b.b_id = sideb.b_id

I currently have separate indices on a_to_b for a_id and b_id and I'm
not sure they're required. Can I delete them without impacting FK
lookups, cacades and joins?

Thanks,

Brian