Fix parsing of UMI tools dedup output to MultiQC about scnanoseq

[lianov]

As a note, our path here for first release will be to disable the outputs from UMI-tools from being passed on to MultiQC at this time. The pipeline already produces samtools stats from before and after dedub, and although this does not provide a bar plot for the % of reads that are detected from UMI-tools, once can see how many reads are dropped from dedup from the already provided metrics (at the top of the MultiQC report).

We will leave this issue open for re-consideration in the future (and the files from UMI-tools are still produced, simply not added to the scnanoseq "out of the box" MultiQC report)