Single-cell type-specific expression analysis.

Lung cancer (LC) is the leading cause of cancer-related mortality worldwide, accounting for millions of deaths annually. Its major subtypes—lung squamous carcinoma (LUSC), lung adenocarcinoma, and small-cell LC—exhibit distinct risk factors and genetic susceptibilities, necessitating the use of subtype-specific biomarkers. Two-sample Mendelian randomization (MR) analyses were conducted using protein quantitative trait loci from the UK Biobank Pharma Proteomics Project and deCODE datasets. A robust analytical framework, including reverse MR, meta-analysis, summary-data-based MR tests, and colocalization, cisMR-cML, MR.CUE and phenotype scanning analyses were used to identify proteins associated with LC risk. We conducted a systematic review to contextualize our research findings. Follow-up analyses, including pathway enrichment, protein-protein interaction network analysis, and druggability evaluations, were used to explore the mechanisms and therapeutic potential of the identified proteins. Significant proteins were validated using population-level proteomic data from the UK Biobank (UKB). The results showed that twenty-five proteins were significantly associated with LC or its subtypes, including 15 novel findings. 60S ribosomal protein L14 (RPL14) and advanced glycosylation end-product-specific receptor (AGER) emerged as the strongest discovery, demonstrating consistent and significant associations across both MR and population-level analyses. RPL14 exhibited positive associations with overall LC risk (MR_meta: odds ratio [OR]: 2.012, 95% confidence interval [CI]: 1.297–3.119; UKB: OR: 1.509, 95% CI: 1.015–2.244). Similarly, AGER showed significant protective effects against LUSC risk (MR_meta: OR: 0.572, 95%CI: 0.368–0.889; UKB: OR: 0.366, 95% CI: 0.158–0.850). Pathway analysis revealed the involvement of these proteins in immune regulation and tumorigenesis. Among the 13 identified druggable targets, RPL14 and AGER showed therapeutic potential as approved or investigational drugs targeting these proteins. These findings offer new insights into the pathogenesis of LC and potential therapeutic targets.肺癌（Lung cancer, LC）是全球范围内癌症相关死亡的首要病因，每年造成数百万例死亡。其主要亚型包括肺鳞状细胞癌（lung squamous carcinoma, LUSC）、肺腺癌与小细胞肺癌，这些亚型具有截然不同的危险因素与遗传易感性，因此亟需开发亚型特异性生物标志物。本研究采用英国生物银行药物蛋白质组学项目与deCODE数据集的蛋白质数量性状位点数据，开展双样本孟德尔随机化（Mendelian randomization, MR）分析。通过构建包含反向孟德尔随机化、荟萃分析、基于汇总数据的孟德尔随机化检验、共定位分析、cisMR-cML、MR.CUE及表型扫描分析的稳健分析框架，筛选与肺癌风险相关的蛋白质；同时开展系统综述以佐证本研究结果。后续通过通路富集分析、蛋白质相互作用网络分析与成药性评估，探索筛选出的蛋白质的作用机制与治疗潜力，并利用英国生物银行（UKB）的群体水平蛋白质组学数据对显著关联的蛋白质进行验证。研究结果显示，共有25种蛋白质与肺癌或其亚型存在显著关联，其中15项为全新发现。60S核糖体蛋白L14（RPL14）与晚期糖基化终产物特异性受体（AGER）为本研究关联性最强的靶点，二者在孟德尔随机化分析与群体水平验证分析中均表现出一致且显著的关联：RPL14与总体肺癌风险呈正相关（孟德尔随机化荟萃分析：比值比[OR]=2.012，95%置信区间[CI]=1.297~3.119；英国生物银行验证队列：OR=1.509，95%CI=1.015~2.244）；类似地，AGER对肺鳞状细胞癌风险具有显著保护作用（孟德尔随机化荟萃分析：OR=0.572，95%CI=0.368~0.889；英国生物银行验证队列：OR=0.366，95%CI=0.158~0.850）。通路富集分析显示，上述蛋白质参与免疫调控与肿瘤发生过程。在13个已识别的成药靶点中，RPL14与AGER可作为已获批或在研药物的靶点，具备明确治疗潜力。本研究结果为肺癌的发病机制提供了全新见解，并提出了潜在治疗靶点。