umi deduplication and sequence length

[lisagrigoreva]

Hello,
Thank you for creating such a great tool! I was wondering whether the UMI deduplication process depends on the read length. For example, if read 1 has a length of 45 bp and read 2 has a length of 75 bp, and they both share the same UMI, will they be collapsed? If so, which length will be preserved in the resulting sequence?

After reading your publication, I have a question: is the hash function applied to the sequence + UMI or to the UMI separately? In other words, does it compare and allow mismatches in the sequence + UMI or only in the UMI?