last-train fastq input

Hello,

When mapping FASTQ files usually there is some preprocessing before mapping. Mainly adapter sequence removal. I am also clustering the reads based on the sequence for a better compression and a slightly faster mapping using clumpify from BBMap.

Hence my problem: does sequence clustering interfere with last-train, so I should use the pre-clustered adapter-free FASTQ or it should not matter?

Many thanks

Darek Kedra