Following the suggestion by the author, I will try to use raw reads instead of trimmed reads.
I meet the same problem because my data have been trimmed by fastp.
