皮纹 43 位点与精神分裂/高血压 GWAS 的全基因组共定位筛查:一份可执行的双方法实验方案

Written by

in

Language

Evidence level: B

皮纹 GWAS 已给出 43 个位点,精神分裂与高血压 GWAS 给出数百个位点——把三张公开汇总统计表放到 coloc 与 eCAVIAR 的贝叶斯框架下逐区间对撞,就能裁决”皮纹-疾病关联是真共享遗传开关,还是连锁不平衡制造的巧合”。

正文

皮纹学与疾病的关系在文献里长期”忽左忽右”:精神分裂症 meta 分析反复报告患者 a-b 脊线数显著降低,高血压研究则发现斗型纹频率与总脊线数偏高;可双生子研究又报告两者无关。同一个指标结论不稳定,通常有两种解释:要么关联是真实但微弱的发育印记,要么纯粹是样本与混杂制造的噪声。过去没有第三种裁决手段,但 2022 年《细胞》的里程碑研究改变了局面——它对 2.3 万余名个体的全基因组扫描与跨族裔荟萃分析,识别出 43 个指纹花纹相关位点(PMID 34995520),并证实指纹相关基因显著富集于肢体发育通路、EVI1 经小鼠功能验证。皮纹第一次有了自己的 GWAS 汇总统计量。同时,精神分裂 GWAS 已定位 287 个位点(PMID 35396580),血压 GWAS 已定位 535+ 个位点(PMID 30224653)。三张全基因组汇总表全部公开可查,意味着”皮纹位点×疾病位点”的全基因组对撞筛查,第一次在数据层面成为可能。

这一筛查的统计工具是共定位分析(colocalization)。其核心思想:如果皮纹 GWAS 与疾病 GWAS 在同一基因组区域内的峰值信号共享同一个因果变异,就说明两个表型由同一个遗传开关驱动;如果只是两个独立的因果变异落在同一个连锁不平衡(LD)区间内,就会造成”看似相关实则无关”的假象。方法学已经成熟:coloc 包的贝叶斯框架(PMID 24830394)输出五个后验概率,其中 PP.H4 代表”两信号共享因果变异”,是判定共定位的核心指标;eCAVIAR(PMID 27866706)则突破单因果变异假设,能处理位点内多个因果变异并存的情况。两者互补:coloc 对”共享单一因果变异”敏感,eCAVIAR 对”多信号位点”稳健。

由此给出可执行的筛查方案。第一步,数据准备:皮纹侧采用 43 位点 GWAS 的汇总统计量(已公开于 NODE 数据库 OEP000198);疾病侧采用精神分裂 PGC 2022 版(287 位点)与血压 GWAS(535+ 位点)汇总统计量。第二步,区间对齐:以皮纹 43 个位点为中心,各取 ±500kb(或 LD 扩展)窗口,在疾病 GWAS 中提取同窗口的全部 SNP 汇总数据;注意东亚人群与欧洲人群 LD 结构不同,需按参考人群分别对齐。第三步,双方法交叉验证:对每个窗口同时跑 coloc 与 eCAVIAR,预设判定标准——PP.H4 ≥ 0.75(coloc)且 eCAVIAR 后验概率 ≥ 0.01(对应其常用阈值),双方法一致命中才算”确认共定位”;单方法命中记为”待复核”。第四步,跨族裔重复:皮纹位点多在东亚人群发现,需用东亚血压 GWAS(PMID 30487518)与欧洲血压 GWAS 分别重复,检验共定位信号是否跨族裔稳定——这一步骤同时是”皮纹-高血压关联是否成立”的终审。

预期结果可按先验分三档。精神分裂侧命中率应最高:DISC1-rs999710 已在家族研究中与皮纹偏离标记共传递(PMID 41037940),家族证据铺垫了共享遗传通路的先验,若共定位命中,等于把家族研究升级为人群水平的因果证据。高血压侧是关键裁决场:文献方向矛盾(斗型纹-血压正相关的病例对照 vs 双生子阴性),若共定位命中则”真共享”成立、矛盾由发育印记强度微弱解释;若完全不命中,则此前关联大概率是混杂噪声。EVI1 位点单独加验:EVI1 已知与白血病相关,斗型纹频率又与白血病易感相关,此位点的共定位结果可顺带检验”皮纹-血液系统疾病”的完整多效性链条。

需要强调的是判定边界。共定位证明”共享因果变异”,不等于”共享命运”——发育基因在胚胎期同时参与皮纹与神经、血管成型,出生后皮纹定型,疾病却还要叠加几十年环境与随机性;即使 PP.H4 全命中,皮纹也只是微弱发育印记,绝非”看手相算命”。方案同时内置三重防误报:双方法一致阈值、跨族裔重复、以及 COLOCdb 等数据库交叉索引(旧文 PMID 37941154 已整理)。这套方案的全部输入数据均已公开,唯一成本是计算资源与统计实现——它是目前把”掌纹藏信息”的古老直觉翻译成可证伪科学命题的最短路径。

证据与局限

已验证: 指纹 GWAS 识别 43 位点并富集肢体发育通路(PMID 34995520);精神分裂 287 位点(PMID 35396580);血压 535+ 位点(PMID 30224653);coloc 贝叶斯方法(PMID 24830394)与 eCAVIAR 多因果变异方法(PMID 27866706)均经模拟与实证检验;DISC1-rs999710 与皮纹偏离家族共传递(PMID 41037940);跨族裔血压位点存在祖先特异变异(PMID 30487518)。

部分验证: 皮纹-精神分裂关联经多项研究确认为稳定的神经发育标记但效应量小;皮纹-高血压关联方向矛盾(病例对照正相关 vs 双生子阴性);43 位点中与疾病位点实际共定位的数量未知。

待验证: 全基因组共定位筛查本身尚无大样本直接检验(本方案即其执行设计);EVI1-白血病多效性通道待共定位+孟德尔随机化双法验证;共定位信号的跨族裔稳定性未知;东亚 vs 欧洲 LD 结构差异对 PP.H4 的影响需敏感性分析。

相关阅读

证据等级: B
参考文献: PMID 34995520, PMID 24830394, PMID 27866706, PMID 35396580, PMID 30224653, PMID 30487518, PMID 41037940


免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。

Genome-wide colocalization screening of skin ridge polymorphism at locus 43 with schizophrenia and hypertension GWAS: a feasible dual-method experimental protocol

Evidence level: B

Dermatoglyphics GWAS has identified 43 loci; schizophrenia and hypertension GWAS have identified hundreds. By placing three public summary statistics into the Bayesian frameworks of coloc and eCAVIAR and conducting interval-by-interval cross-validation, one can adjudicate whether “dermatoglyphics-disease associations represent true shared genetic switches or coincidental linkage disequilibrium artifacts.”

Main Text

The relationship between dermatoglyphics and disease in the literature has long been inconsistent: schizophrenia meta-analyses repeatedly report a significant reduction in ridge counts along the a-b lines, while hypertension studies find higher frequencies of whorls and total ridge counts; yet twin studies report no association. The instability of conclusions regarding the same metric typically stems from two explanations: either the association reflects a real but weak developmental imprint, or it is purely noise generated by sample bias and confounding. Previously, there was no third adjudication tool; however, the landmark 2022 Cell study altered this landscape. Through a whole-genome scan and cross-ethnic meta-analysis of over 23,000 individuals, it identified 43 fingerprint-pattern-associated loci (PMID 34995520) and confirmed that fingerprint-related genes are significantly enriched in limb development pathways, with EVI1 validated via mouse function. Dermatoglyphics finally possesses its own GWAS summary statistics. Simultaneously, schizophrenia GWAS has localized 287 loci (PMID 35396580), and blood pressure GWAS has localized 535+ loci (PMID 30224653). With all three whole-genome summary statistics publicly available, a genome-wide cross-validation screening of “dermatoglyphics loci × disease loci” has become feasible at the data level for the first time.

The statistical tool for this screening is colocalization analysis. Its core premise is as follows: if the peak signals in dermatoglyphics GWAS and disease GWAS within the same genomic region share a single causal variant, it indicates that both phenotypes are driven by the same genetic switch; if two independent causal variants merely fall within the same linkage disequilibrium (LD) interval, it creates an illusion of “apparent association but no true causality.” The methodology is mature: the Bayesian framework of the coloc package (PMID 24830394) outputs five posterior probabilities, among which PP.H4 represents “two signals sharing a causal variant” and serves as the core metric for determining colocalization; eCAVIAR (PMID 27866706) breaks the single-causal-variant assumption and can handle cases where multiple causal variants coexist within a locus. They are complementary: coloc is sensitive to “shared single causal variants,” while eCAVIAR is robust for “multi-signal loci.”

Based on this, we propose an executable screening protocol. Step 1: Data Preparation: On the dermatoglyphics side, use summary statistics from the 43-loci GWAS (publicly available in NODE database OEP000198); on the disease side, use summary statistics from the schizophrenia PGC 2022 version (287 loci) and blood pressure GWAS (535+ loci). Step 2: Interval Alignment: Centering on the 43 dermatoglyphics loci, extract all SNP summary data within ±500kb (or LD-expanded) windows from the disease GWAS; note that LD structures differ between East Asian and European populations, requiring alignment by reference population. Step 3: Dual-Method Cross-Validation: Run both coloc and eCAVIAR for each window, presetting decision criteria—PP.H4 ≥ 0.75 (coloc) and eCAVIAR posterior probability ≥ 0.01 (corresponding to its common threshold)—where only consistent hits from both methods count as “confirmed colocalization”; single-method hits are recorded as “pending verification.” Step 4: Cross-Ethnic Replication: Since dermatoglyphics loci are predominantly discovered in East Asian populations, repeat using East Asian blood pressure GWAS (PMID 30487518) and European blood pressure GWAS separately to test whether colocalization signals remain stable across ethnicities—this step simultaneously serves as the final audit for “whether the dermatoglyphics-hypertension association holds.”

Expected results can be categorized into three tiers based on priors. The schizophrenia side should yield the highest hit rate: DISC1-rs999710 has already shown co-transmission with dermatoglyphics deviation markers in family studies (PMID 41037940); family evidence provides prior support for shared genetic pathways, so a colocalization hit would upgrade family studies to population-level causal evidence. The hypertension side is the key adjudication arena: literature directions are contradictory (case-control positive correlation for whorls-bp vs. twin negative), so if colocalization hits, “true sharing” is established and contradictions explained by weak developmental imprint intensity; if completely missed, prior associations are likely confounding noise. EVI1 locus specific validation: EVI1 is known to be associated with leukemia, and whorl frequency is also associated with leukemia susceptibility; thus, the colocalization result for this locus can concurrently validate the complete pleiotropy chain of “dermatoglyphics-hematological disease.”

It is crucial to emphasize the decision boundaries. Colocalization proves “shared causal variants,” not “shared fate”—developmental genes participate in fingerprint, neural, and vascular formation simultaneously during embryogenesis; fingerprints are fixed after birth, while diseases still accumulate decades of environmental and random factors; even if PP.H4 hits fully, dermatoglyphics represents only a weak developmental imprint, far from “fortune-telling via palm reading.” The protocol also embeds three anti-false-positive safeguards: dual-method consistent thresholds, cross-ethnic replication, and cross-indexing with databases like COLOCdb (previous work PMID 37941154 already organized). All input data for this protocol are publicly available; the only cost is computational resources and statistical implementation—it represents the shortest path to translating the ancient intuition that “palm lines hide information” into a falsifiable scientific proposition.

Evidence and Limitations

Verified: Fingerprint GWAS identified 43 loci enriched in limb development pathways (PMID 34995520); schizophrenia 287 loci (PMID 35396580); blood pressure 535+ loci (PMID 30224653); coloc Bayesian method (PMID 24830394) and eCAVIAR multi-causal variant method (PMID 27866706) both validated via simulation and empirical testing; DISC1-rs999710 co-transmitted with dermatoglyphics deviation in families (PMID 41037940); cross-ethnic blood pressure loci exhibit ancestry-specific variants (PMID 30487518).

Partially Verified: Dermatoglyphics-schizophrenia association confirmed as a stable neurodevelopmental marker but with small effect sizes across multiple studies; dermatoglyphics-hypertension association shows contradictory directions (case-control positive vs. twin negative); the actual number of colocalized loci among the 43 is unknown.

Pending Verification: Whole-genome colocalization screening itself lacks large-sample direct validation (this proposal is its execution design); EVI1-leukemia pleiotropy channel requires dual validation via colocalization + Mendelian randomization; cross-ethnic stability of colocalization signals is unknown; impact of East Asian vs. European LD structural differences on PP.H4 requires sensitivity analysis.

Related Reading

Evidence Level: B
References: PMID 34995520, PMID 24830394, PMID 27866706, PMID 35396580, PMID 30224653, PMID 30487518, PMID 41037940


Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *