A multigenic approach to evaluating prostate cancer risk in a systematic replication study
- 作者：Fang-Chi Hsu ; Sara Lindströ ; m ; Jielin Sun ; Fredrik Wiklund ; Shyh-Huei Chen ; Hans-Olov Adami ; Aubrey R. Turner ; Wennuan Liu ; Katarina Bä ; lter ; Jin Woo Kim ; Pä ; r Stattin ; Bao-li Chang ; Willi
- 关键词：Multistability ; Multiperiodicity ; Cohen– ; Grossberg neural networks ; Exponential stability ; Exponentially attractive ; Saturation regions
- 刊名：Cancer Genetics and Cytogenetics
- 出版年：2008
- 期刊代码：43_01654608
- 类别：med
- 出版时间：June 2008
- 卷：183
- 期：2
- 页码：94-98
- 文件大小：237 K
摘要
Although it is well known that multiple genes may influence prostate cancer risk, most current efforts at identifying prostate cancer risk variants rely on single-gene approaches. In previous work using mostly single-gene approaches, we observed significant associations (P < 0.05) for 6 of 46 polymorphisms in five genes in a Swedish prostate cancer case-control study population. We now report on the higher-order gene–gene interactions among those 46 genetic variants and the combined effect of the six polymorphisms with significant main effects for association with prostate cancer risk in 795 controls and 1,461 cases. Classification and regression tree analysis was used to evaluate higher-order gene–gene interactions. No interactions were confirmed by the result from logistic regressions. For the combined analysis, we tested the hypothesis that individuals carrying multiple copies of risk variants are at increased risk for prostate cancer. Individuals carrying more than eight copies of any risk variant were almost twofold more likely to get prostate cancer (OR = 1.99, P = 0.0014). A significant trend relationship was observed (P < 0.0001). In the present study, additive effects but not multiplicative effects among these six polymorphisms with significant main effects were observed.