Computational protocol: Association between Genetic Variants in DNA Double-Strand Break Repair Pathways and Risk of Radiation Therapy-Induced Pneumonitis and Esophagitis in Non-Small Cell Lung Cancer

Similar protocols

Protocol publication

[…] A Chi-square test and a Student’s t-test were used to assess the distribution of covariates between patients with or without an event. Multivariate logistic regression was used to estimate the main effect of single SNP on risk of developing toxicities adjusted for age, sex, pack year, clinical stage, performance status, chemoradiotherapy, radiation treatment type, forced expiratory volume in the first second (FEV1), carbon monoxide diffusing capacity (DLCO) percentage, planning tumor volume (PTV), mean dose (mean esophagus dose for esophagitis and mean lung dose for pneumonitis). For each SNP dominant, recessive and additive models were analyzed, and only the best-fitting model were reported. In order to expand the coverage of the validation panel, we also included proxy SNPs that are in high linkage disequilibrium (LD > 80%) with the original genotyped SNP from the discovery phase and analyzed in the validation population. The proxy SNPs were identified using SNAP [] based on LD information calculated using phased genotype data from the International HapMap Project and the 1000 Genomes Project.If the proxy SNPs showed a similar effect to the original SNP in the discovery phase, meta-analysis was performed to summarize the effects from discovery and validation populations. Heterogeneity was estimated using χ2-based Q-statistics. A fixed effect model was used when heterogeneity was absent (p for heterogeneity >0.05). An unfavorable genotype (UFG) was defined as the allele associated with an increased risk of developing toxicities. A joint analysis all UFGs for each patient and risk of esophagitis or pneumonitis was conducted with subjects stratified by level of risk in tertiles. All statistical analyses were two-sided. The analysis described above was performed using STATA software (version 10, STATA Corp, College Station, TX, USA).HaploReg [] was used for functional annotations of candidate SNPs. PolyPhen-2 [], SNPeffect [], SIFT [] and SNPs3D [] were used to predict the function of missense variant on protein function. VEGAS was used to perform gene-based tests, which produces a gene-based test statistic based on a simulation to calculate an empirical gene-based p-value []. […]

Pipeline specifications

Software tools HaploReg, PolyPhen, SIFT, SNPs3D, VEGAS
Databases SNPeffect
Application GWAS
Organisms Homo sapiens
Diseases Carcinoma, Non-Small-Cell Lung, Esophagitis, Pneumonia, Radiation Injuries, Drug-Related Side Effects and Adverse Reactions