通路富集

F富 · 旁支 · 图 1 张 · 发表版表格 2 个　（流程见 v5 定稿）

挂在 4X 之后的虚线旁支，什么都不否决。★ 必须按 MHC 分组看。

---

图

Figure F-enrich. Functional enrichment of the 28 proteins still retained after the directionality and artifact checks, analysed as a whole and split by MHC membership

完整图注（英文，与投稿版一致）

Figure F-enrich. Functional enrichment of the 28 proteins still retained after the directionality and artifact checks, analysed as a whole and split by MHC membership.

Gene sets. All retained proteins, n = 28; of these 11 are encoded in the MHC region and 17 are not. The two subsets are disjoint and exhaust the full set. Numbers to the right of each bar are the count of query genes in that term over the number of query genes that could be mapped in that annotation source; the denominator varies between sources because not every gene is annotated in every source.

The reason for splitting. Taken as one set, the retained proteins look like an immune signature: 6 of the 9 enriched terms in the combined analysis are immune-related. Splitting shows where that comes from. The 11 MHC-region proteins return 11 terms, 8 of them immune-related. The 17 non-MHC proteins return 5 terms, 0 of them immune-related; those terms are all vesicle-related cellular components. The immune signature of the combined set is therefore carried by the MHC block, and reporting it for the whole set would misattribute it to the study's candidate proteins in general.

Interpret term counts with care. Gene Ontology terms are hierarchical and the terms returned within each panel are largely nested: in the non-MHC panel, vesicle contains intracellular vesicle which contains cytoplasmic vesicle, and vesicle membrane contains cytoplasmic vesicle membrane. Five terms there represent essentially one finding, not five independent ones. The same applies to the immune terms in the MHC panel.

Method: g:Profiler with the g:SCS multiple-testing correction, run separately for each gene set against GO:BP, GO:CC, GO:MF, KEGG and Reactome. The dashed line marks adjusted p = 0.05. Per-source domain sizes and significance thresholds are given in the supplementary table pathway_enrichment_sources.csv. All significant terms are shown; nothing is truncated.

Role in the analysis pipeline: enrichment is a descriptive side-branch. It does not filter, rank, veto or promote any protein, and no statement elsewhere in this study depends on it. It is reported because the MHC contribution is worth making explicit, not because it contributes evidence about any individual candidate.

Data sources and sample sizes. Exposure: UKB-PPP plasma proteome (Olink Explore 3072), 34,557 European participants, 1,954 proteins with a cis-pQTL. Outcomes: FinnGen R9 -- Diabetic retinopathy 10,413 cases / 308,633 controls; Diabetic maculopathy 3,572 cases / 308,547 controls; Diabetic nephropathy 4,111 cases / 308,539 controls; Diabetic neuropathy 2,843 cases / 271,817 controls.

---

发表版表格

文件	下载
pathway_enrichment.csv	下载
pathway_enrichment_sources.csv	下载

本页图为网页版 PNG（最宽 1600 px）。投稿用的矢量 PDF 与 600 dpi TIFF 体积较大，留在仓库 results/enrichment/figures/，不随文档站分发。