三阴性乳腺癌具有侵袭性高、易转移、耐药性强、预后差的特点，目前治疗手段获益不足。因此对其耐药标记物和机理研究迫在眉睫。前期实验中，我们已使用压力循环-卫星扫描技术（PCT-SWATH）与机器学习结合，找到39个三阴性乳腺癌和37种腔面型乳腺癌细胞系差异表达的24个蛋白，并结合这些细胞对90个抗乳腺癌药物的反应和基因组和转录组数据，利用机器学习初探药物敏感性相关蛋白。我们进一步选择如紫杉醇等多种治疗三阴性乳腺癌临床一线药物处理5种三阴性和4种腔面型乳腺癌细胞系作为模型，对比时序性给药研究建立药物蛋白动态互作网络，结合生物信息学，发现三阴性乳腺癌耐药的生物标记物甚至靶点。继而利用细胞水平、PDX模型探究耐药机理，最后在临床队列中使用靶向蛋白质组技术验证实际临床价值，完成对三阴性乳腺癌耐药的系统研究，为三阴性乳腺癌患者耐药的早期诊断及临床治疗新靶点研究提供新的理论依据。

Triple negative breast cancers (TNBCs) are highly invasive, metastatic, resistant to most drugs, leading to poor prognosis. The needs to study biomarkers for its drug resistance and overcome the resistance are pressing. Pressure cycling technology coupled with SWATH mass spectrometry (PCT-SWATH) enables high-throughput quantitative analysis of proteome of small amount of clinical samples with high degree of accuracy, generating proteomic big data, and allows systematic investigation of drug resistance. In pilot studies, we have analyzed the proteomes of 39 TNBCs and 37 luminal breast cancer cell lines as controls using PCT-SWATH. Through machine learning we narrowed our focus to 24 proteins which are specifically expressed in TNBCs. We will further study the TNBCs cells upon treatment of clinically used drugs using PCT-SWATH, to identify protein biomarkers predicting drug resistance. Next we will investigate the drug resistance mechanisms using cell and mouse models, followed by verification of drug resistance biomarkers using Parallel Reaction Monitoring (PRM) targeted proteomics technology in a breast cancer cohort, and then explore circumventing the drug resistance by modulating these druggable biomarkers.