Contigs have different mutations than raw reads

Question

[Rashedul Islam]

Hi,

I ran two assembly softwares and called mutations from the contigs. However, when I compared the contigs with raw reads, I found some instances where raw reads does not have mutations but the contigs have. I don't know how to explain this. Please see the attached.

[Friederike]

Which assemblers did you run and how?

[Rashedul Islam]

The first one is megahit ($megahit -1 fastq1 -2 fastq2 -o /path/) and second one is metaspades ($metaspades.py -1 fastq1 -2 fastq2 -o /path/). Reads were trimmed with Trimmomatic. Its an RNA virus genome.