How to process sequence data sets with different length from different platforms?

I have three sets of sequence data, including single-end sequence, paired-end sequence (300bp X 2) and paired-end sequence (250bp X 2). I wander how to generate ASV from these data efficiently and correctly? Thanks