To reduce any technical variation between plates and studies Olink recommend that normalization is performed, as this will reduce the technical variation. If samples can be considered randomized (read more about randomization here ).
Olink will use intensity normalization by applying median centering normalization between plates for each assay. This will effectively reduce any shift in median signal between plates. However, if samples are not randomized, e.g. in regards of study groups or other variables, true biological differences may be removed if intensity normalization is used.
When randomization is not feasible between plates or studies reference samples must be included on all plates. Olink recommend that 8-16 reference samples are chosen and included on each plate or in each study, which can be utilized to normalize and bridge between plates and studies. If possible, these reference samples should be representative for the entire study, for example include all study groups, and should be matrix-matched with other samples.
See the related FAQs:
For more information see our white paper, “Data normalization and standardization“.