Problems with QIAseq spike-in normalization

Question

[lluc.cabus]

Hi,

I'm trying to use the QIAseq miRNA library QC spike-ins to normalize a set of samples. This kit contains several spike-in miRNAs at 3 different concentrations: 0.75fM, 7.5fM and 37.5fM. The problem I am facing is that several miRNAs that are put in the lowest of medium concentrations are found in higher quantities than the miRNAs that are put in higher the highest concentration (I attach here a plot of the correlation between counts of spike-in and its concentration).

The problem here is: can I use the spike-ins here to normalize the samples if some of the spike-ins with higher concentration have lower counts than the spike-ins with lower concentrations?

Thanks, Lluc!