Collapsing my paired-end reads before or after mapping ?

Hi everyone,

I am about to map some Paired-End (PE) reads (aDNA) and I don't know what is the pro and cons about those two methods:

1. Run fastp (to remove the left-over adapters) -> map my PE dataset (using BWA) -> single BAM -> remove duplicates using Picard -> Mapping Quality filtering
2. Run Clip&Merge from EAGER (that collapse reads and trim adapters, specialised for aDNA) -> map my new SE dataset (using BWA) -> remove duplicates using Picard -> Mapping Quality filtering

My question is; is there any advantage for any of those two methods ?

Thanks a lot.