Plasma proteomics in two Swedish cohorts with Ovarian Cancer

The plasma proteome was analyzed using the proximity extension assay (PEA) as implemented in the Explore HT-version (Olink Proteomics AB, Uppsala, Sweden). The samples were analyzed at the Olink Service Laboratory in Uppsala, Sweden. In brief, the PEA is based on pairs of antibodies equipped with probes, DNA single-strand oligonucleotide reporter molecules, that bind to their respective target if present. Target binding by both probes in close proximity generates double-stranded DNA amplicons which are then quantified by next generation sequencing. Here, 5416 unique proteins were characterized in each of the 404 samples. The samples were randomized across plates with respect to cohort and diagnose (benign or malignant). In the resulting data-file provided by the analysis platform, each individual protein measurement, assay and sample is labelled depending on passing or failing quality control as provided by the instrument software. Here, a total on 768 individual measurements (corresponding to 0.036%, 768/(404*5416)) did not pass quality control and were removed from further analyses. Two proteins (apolipoprotein E (APOE) and fibronektin 1 (FN1)) had detection rates below 95% across all samples and were removed from further analyses. The resulting NPX values are on a log2 scale and in the logarithmic phase of the curve, one (1) increase of the NPX value corresponds to a doubling of the protein content. In the resulting data, a high NPX value corresponds to a high protein concentration. Each of the measured proteins has a lower limit of detection (LOD) given in the same NPX-scale which is determined at run time. Here, measurements under LOD were kept as is in the downstream analysis. After quality control, the detection rate across all samples were 99.5 to 99.8%.