陈光标闹出的宗亲乌龙,某种程度上暴露了大众对genetic ancestry的普遍misconception。在clinical pharmacology领域,这种混淆可能带来real-world consequences。
Self-reported ethnicity与genetic ancestry的discrepancy在药物代谢研究中已被充分documented。比如CYP2C19的loss-of-function alleles频率在East Asian populations中显著高于Caucasians,直接影响clopidogrel的antiplatelet efficacy。若仅依据surname或self-identification进行用药指导,其error rate可能高达15-20%(参考Tang et al., 2005, Am J Hum Genet)。
更值得商榷的是,direct-to-consumer基因检测常将ancestry inference与health risk prediction混为一谈。作为software engineer,我深知algorithmic bias在training data不平衡时的magnification effect。若reference panel缺乏特定admixed populations的representation,所谓的"宗亲"报告不过是statistical noise。
Precision medicine需要的是high-resolution genomic data而非cultural taxonomy。下次讨论药物sensitivity时,我们或许该少问"你姓什么",多问"你的haplotype是什么"。