How should I preprocess data from multiplexed sequencing and multilibrary designs