Table 2.
Significant differentially abundant proteins identified
| GI# | Protein description | Upregulated in | Specificity | Sensitivity | P value (Fisher's exact, two sided) |
| 4885165 | Cystatin A (stefin A) | Control | 0.650 | 1.000 | 1.92 × e-08 |
| 6995994 | Aggrecan 1 (chondroitin sulfate proteoglycan 1) | Control | 0.650 | 0.974 | 2.30 × e-07 |
| 1651921 | Dermcidin | Control | 0.600 | 1.000 | 1.13 × e-07 |
| 4502027 | Albumin | OA | 0.950 | 0.718 | 7.96 × e-07 |
| 4502067 | α1-Microglobulin/bikunin precursor | OA | 0.950 | 0.718 | 7.96 × e-07 |
| 4503689 | Fibrinogen, α chain isoform α-E preprotein | OA | 0.950 | 0.718 | 7.96 × e-07 |
| 4503715 | Fibrinogen, γ chain isoform γ-A precursor | OA | 1.000 | 0.744 | 1.43 × e-08 |
| 4557225 | α2-Macroglobulin | OA | 0.950 | 0.718 | 7.96 × e-07 |
| 4557325 | Apolipoprotein E | OA | 1.000 | 0.744 | 1.43 × e-08 |
| 4557327 | Apolipoprotein H (β2-glycoprotein I) | OA | 1.000 | 0.744 | 1.43 × e-08 |
| 4557385 | Complement component 3 (gel slice 3) | OA | 0.950 | 0.718 | 7.96 × e-07 |
| 4557385 | Complement component 3 (gel slice 5) | OA | 1.000 | 0.744 | 1.43 × e-08 |
| 4557485 | Ceruloplasmin (ferroxidase) | OA | 0.950 | 0.718 | 7.96 × e-07 |
| 4826762 | Haptoglobin | OA | 0.950 | 0.718 | 7.96 × e-07 |
| 9257232 | Orosomucoid 1 | OA | 0.850 | 0.667 | 2.51 × e-04 |
| 32483410 | Group specific component (vitamin D binding protein) | OA | 1.000 | 0.744 | 1.43 × e-08 |
| 50345296 | Complement component 4B preprotein (NP_001002029) | OA | 1.000 | 0.744 | 1.43 × e-08 |
| 5147611 | PREDICTED: similar to apolipoprotein A-1 precursor (apo-A-1; XP_496536) | OA | 0.950 | 0.718 | 7.96 × e-07 |
| 55743122 | Retinol-binding protein 4, plasma precursor (NP_006735) | OA | 0.900 | 0.692 | 1.87 × e-05 |
Eighteen proteins that were significantly differentially abundant (protein area) across control (n = 20) and OA (n = 39) groups by Wilcoxon's ranksum test at P < 1 × e-06. Sensitivity/specificity for each protein are calculated with respect to the number of samples of each group having protein area above or below the median area across all samples. The significance of median area dichotomy and true group label is assessed by two-sided Fisher's exact test. GI#, GenInfo accession.