Supporting figures and tables.

Figure S1, One polymorphic site is removed at a time from all the genotypic sets in the population to predict the probable risk allele. The effect is studied in terms of the number of unique cases with precancer (leukoplakia) and cancer remaining after each locus deletion. The number under each bar in the X axis represents the omitted locus. A) Exclusion of SNP2 shows highest decline in the size of population suffering from cancer. B) Deletion of any locus is not associated with increase or decrease in precancerous population. C) Exclusion of SNP4 shows highest decline in the size of healthy (control) population. Figure S2, Restructuring of the case-control specific genotypes as different supersets are created taking 4 SNPs at a given time, the one removed each time is denoted by "*". Figure S3, One SNP at a time is removed from all the genotypic set in the ACS-control population to observe the effect after omission of one locus. The removed locus is denoted by * in the genotype supersets taking 4 loci at a time. The effect is studied in terms of the distribution of population under different conditions namely Case, Control and Common groups. Table S1, Polymorphisms identified in P2RY1 and P2RY12 genes. Table S2, A. Frequency of combination of genotypes among ACS patients, respective controls and combined individuals. B. Frequency of combination of genotypes among oral cancer patients, precancer patients and controls. Table S3, p values after omission of one SNP from oral cancer and control population. Omission of SNP2 and SNP4 significantly decrease the case specific and control specific genotypic fraction respectively. Therefore SNP2 might be called as ‘risk SNP’ and SNP4 as ‘protective SNP’. The significant p-values are marked as ‘*’. Table S4, A. Frequency of combination of genotypes obtained after omission of one SNP from ACS and control population B. Frequency of combination of genotypes obtained after omission of one SNP from oral cancer, leukoplakia and control population. Omitted SNPs are marked as ‘*’. (DOC)

图S1：每次从群体中的所有基因型集中移除一个多态性基因座，以预测潜在的风险等位基因。研究效应通过每次删除一个基因座后剩余的伴癌前病变（口腔白斑）和癌症的独特病例数来衡量。X轴每根条形下方的数字代表被移除的基因座。A) 移除SNP2时，癌症受累群体的规模下降幅度最大。B) 移除任意基因座均不会导致癌前病变群体规模出现增减变化。C) 移除SNP4时，健康（对照）人群规模下降幅度最大。图S2：针对病例-对照特异性基因型进行重构，每次选取4个单核苷酸多态性（Single Nucleotide Polymorphism, SNP）构建不同的超集，每次被移除的位点以"*"标注。图S3：每次从急性冠状动脉综合征（Acute Coronary Syndrome, ACS）对照群体的所有基因型集中移除一个SNP，以观察移除单个基因座后的效应。每次选取4个基因座构建基因型超集时，被移除的基因座以"*"标注。研究效应通过病例组、对照组及共分组三类不同条件下的人群分布来衡量。表S1：P2RY1与P2RY12基因中鉴定出的多态性位点。表S2：A. 急性冠状动脉综合征患者、对应对照人群以及合并人群中基因型组合的频率。B. 口腔癌患者、癌前病变患者与对照人群中基因型组合的频率。表S3：从口腔癌与对照人群中移除单个SNP后的P值。移除SNP2与SNP4分别会显著降低病例特异性与对照特异性的基因型占比，因此可将SNP2定义为"风险SNP"，SNP4为"保护性SNP"。具有统计学显著性的P值以"*"标注。表S4：A. 从ACS与对照人群中移除单个SNP后得到的基因型组合频率。B. 从口腔癌、口腔白斑与对照人群中移除单个SNP后得到的基因型组合频率。被移除的SNP以"*"标注。(DOC)