In two way replication, all the tables to be replicated should exists on target before loading data in order to establish starting point for replication. Unlike One way replication where auto-generated datapump script creates tables and load data into it in single execution. Thus, There are two different approches for creating starting point in two way replication. Selection of correct approach depends upon the amount of data and number of tables.
1. Using a Standby Database
If there are many tables to be replicated with large amounts of data, It is recommended to use standby database
2. Using datapump
If you have just a few tables that you want to replicate and not so much data, then it is easier and faster with datapump