Sample design (top panel): Donor samples from 39 rheumatoid arthritis (RA) patients and 29 healthy individuals (control, Ctrl) were block randomized into two batches based on gender and disease classification. Equal aliquots of all serum samples were mixed (serum pool) to prepare three sample replicates for each batch (sample preparation quality controls, SPQCs). An additional aliquot was added to each batch and prepared in parallel with all other samples, but without the addition of ARP (negative control).

Sample preparation using a modified FASP protocol (middle panel): Serum samples containing 2.0 mg of protein were derivatized with ARP, ultra-filtered (10 kDa cutoff) and digested with trypsin. An aliquot (50 µg protein) was taken for analysis and the remaining sample was enriched by avidin affinity chromatography. Aliquots from each fraction were combined as matrix-specific LC-MS run quality controls (RQCs) to monitor instrumental drift. Additional cohort-specific enriched digest pools, the Ctrl pool and the RA pool, were prepared from the enriched fractions.

Data acquisition (bottom panel): RQCs and the enriched Ctrl and RA pools were analyzed by two DDA methods using precursor ion preference lists to generate DDA spectral libraries of ARP peptides. In addition, ARP peptides identified in the ProteomeXchange data set deposited as PXD023738 were cross-referenced based on their iRT. For quantitative interrogation, RQCs, SPQCs, NCs, and individual RA and Ctrl samples were measured using UDMS^E. Initial DDA measurements were used to condition the system, followed by SPQCs from Batch #1 (B1) and Batch #2 (B2). Next, all donor samples from B1 and B2 were measured, interspersed with RQCs every 10 samples.