Dataset Information

Machine learning on genome-wide association studies to predict the risk of radiation-associated contralateral breast cancer in the WECARE Study.

ABSTRACT: The purpose of this study was to identify germline single nucleotide polymorphisms (SNPs) that optimally predict radiation-associated contralateral breast cancer (RCBC) and to provide new biological insights into the carcinogenic process. Fifty-two women with contralateral breast cancer and 153 women with unilateral breast cancer were identified within the Women's Environmental Cancer and Radiation Epidemiology (WECARE) Study who were at increased risk of RCBC because they were ? 40 years of age at first diagnosis of breast cancer and received a scatter radiation dose > 1 Gy to the contralateral breast. A previously reported algorithm, preconditioned random forest regression, was applied to predict the risk of developing RCBC. The resulting model produced an area under the curve (AUC) of 0.62 (p = 0.04) on hold-out validation data. The biological analysis identified the cyclic AMP-mediated signaling and Ephrin-A as significant biological correlates, which were previously shown to influence cell survival after radiation in an ATM-dependent manner. The key connected genes and proteins that are identified in this analysis were previously identified as relevant to breast cancer, radiation response, or both. In summary, machine learning/bioinformatics methods applied to genome-wide genotyping data have great potential to reveal plausible biological correlates associated with the risk of RCBC.

SUBMITTER: Lee S

PROVIDER: S-EPMC7046218 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning on genome-wide association studies to predict the risk of radiation-associated contralateral breast cancer in the WECARE Study.

Lee Sangkyu S Liang Xiaolin X Woods Meghan M Reiner Anne S AS Concannon Patrick P Bernstein Leslie L Lynch Charles F CF Boice John D JD Deasy Joseph O JO Bernstein Jonine L JL Oh Jung Hun JH

PloS one 20200227 2

The purpose of this study was to identify germline single nucleotide polymorphisms (SNPs) that optimally predict radiation-associated contralateral breast cancer (RCBC) and to provide new biological insights into the carcinogenic process. Fifty-two women with contralateral breast cancer and 153 women with unilateral breast cancer were identified within the Women's Environmental Cancer and Radiation Epidemiology (WECARE) Study who were at increased risk of RCBC because they were ≤ 40 years of age ...[more]

PMID: 32106268

Similar Datasets

Project description:Background:Women with unilateral breast cancer (UBC) are at risk of developing a subsequent contralateral breast cancer (CBC). Common variants are associated with breast cancer risk. Whether these influence CBC risk is unknown. Methods:Participants were breast cancer cases from the population-based Women's Environmental Cancer and Radiation Epidemiology (WECARE) Study. Sixty-seven established breast cancer risk loci were genotyped directly or by imputation in 1459 case subjects with CBC and 2126 UBC control subjects. An unweighted polygenic risk score (PRS) was created by summing the number of risk alleles for each directly genotyped single nucleotide polymorphism (SNP), or for imputed loci, the imputed dosage. A weighted PRS was calculated similarly, but where each SNP's contribution was weighted by the published per-allele log odds ratio. Unweighted and weighted polygenic risk scores and CBC risk were modeled using conditional logistic regression. Cumulative CBC risk was estimated and benchmarked using Surveillance, Epidemiology, and End Results population incidence rates. Results:Both unweighted and weighted PRS were statistically significantly associated with CBC risk. The adjusted risk ratio of CBC in women in the upper quartile of unweighted PRS compared with the lowest quartile was 1.63 (95% confidence interval [CI] = 1.33 to 2.00). The estimated 10-year cumulative risk for women in the upper quartile of the unweighted PRS was 7.4% (95% CI?=?6.0% to 9.1%). For women in the upper quartile of the weighted PRS, the risk ratio for CBC was 1.75 (95% CI?=?1.41 to 2.18) compared with women in the lowest quartile. There was no statistically significant heterogeneity by age, treatment (radiation therapy dose, tamoxifen, chemotherapy), estrogen receptor status of the first primary, histology of the first primary, length of at-risk period for CBC, or breast cancer family history. Conclusions:Common genomic variants associated with the development of first primary breast cancer are also associated with the development of CBC; the risk is strongest among those who carry more risk alleles.

Project description:Genome-wide association studies, focusing primarily on unilateral breast cancer, have identified single nucleotide polymorphisms (SNPs) in a number of genomic regions that have alleles associated with a significantly increased risk of breast cancer. In the current study we evaluate the contributions of these previously identified regions to the risk of developing contralateral breast cancer. The most strongly disease-associated SNPs from prior studies were tested for association with contralateral breast cancer. A subset of these SNPs, selected upon their main effects on contralateral breast cancer risk was further evaluated for interaction with treatment modalities and estrogen receptor (ER) status.We genotyped 21 SNPs in 708 women with contralateral breast cancer and 1394 women with unilateral breast cancer who serve as the cases and controls in the Women's Environment, Cancer and Radiation Epidemiology (WECARE) Study. Records of treatment and ER status were available for most of WECARE Study participants. Associations of SNP genotypes and risk for contralateral breast cancer were calculated with multivariable adjusted conditional logistic regression methods.Multiple SNPs in the FGFR2 locus were significantly associated with contralateral breast cancer, including rs1219648 (per allele rate ratio (RR) = 1.25, 95%CI = 1.08-1.45). Statistically significant associations with contralateral breast cancer were also observed at rs7313833, near the PTHLH gene (per allele RR = 1.26, 95%CI = 1.08-1.47), rs13387042 (2q35) (per allele RR = 1.19, 95%CI = 1.02-1.37), rs13281615 (8q24) (per allele RR = 1.21, 95%CI = 1.04-1.40), and rs11235127 near TMEM135 (per allele RR = 1.26, 95%CI = 1.04-1.53). The A allele of rs13387042 (2q35) was significantly associated with contralateral breast cancer in ER negative first tumors while the A allele of rs11235127 (near TMEM135) was significantly associated with contralateral breast cancer in ER positive first tumors. Although some SNP genotypes appeared to modify contralateral breast cancer risk with respect to tamoxifen treatment or particular radiation doses, trend tests for such effects were not significant.Our results indicate that some common risk variants associated with primary breast cancer also increase risk for contralateral breast cancer, and that these risks vary with the ER status of the first tumor.

Project description:BackgroundTamoxifen treatment greatly reduces a woman's risk of developing a second primary breast cancer. There is, however, substantial variability in treatment response, some of which may be attributed to germline genetic variation. CYP2D6 is a key enzyme in the metabolism of tamoxifen to its active metabolites, and variants in this gene have been associated with reduced tamoxifen metabolism. The impact of variation on risk of contralateral breast cancer (CBC) is unknown.MethodsGermline DNA from 1514 CBC cases and 2203 unilateral breast cancer controls was genotyped for seven single nucleotide polymorphisms, one three-nucleotide insertion-deletion, and a full gene deletion. Each variant has an expected impact on enzyme activity, which in combination allows for the classification of women as extensive, intermediate, and poor metabolizers (EM, IM, and PM respectively). Each woman was assigned one of six possible diplotypes and a corresponding CYP2D6 activity score (AS): EM/EM (AS = 2), EM/IM (AS = 1.5), EM/PM (AS = 1), IM/IM (AS = 0.75), IM/PM (AS = 0.5), and PM/PM (AS = 0). We also collapsed categories of the AS to generate an overall phenotype (EM, AS ≥ 1; IM, AS = 0.5-0.75; PM, AS = 0). Rate ratios (RRs) and 95% confidence intervals (CIs) for the association between tamoxifen treatment and risk of CBC in our study population were estimated using conditional logistic regression, stratified by AS.ResultsAmong women with AS ≥ 1 (i.e., EM), tamoxifen treatment was associated with a 20-55% reduced RR of CBC (AS = 2, RR = - 0.81, 95% CI 0.62-1.06; AS = 1.5, RR = 0.45, 95% CI 0.30-0.68; and AS = 1, RR = 0.55, 95% CI 0.40-0.74). Among women with no EM alleles and at least one PM allele (i.e., IM and PM), tamoxifen did not appear to impact the RR of CBC in this population (AS = 0.5, RR = 1.08, 95% CI 0.59-1.96; and AS = 0, RR = 1.17, 95% CI 0.58-2.35) (p for homogeneity = - 0.02).ConclusionThis study suggests that the CYP2D6 phenotype may contribute to some of the observed variability in the impact of tamoxifen treatment for a first breast cancer on risk of developing CBC.

Project description:Staphylococcus aureus is a major human and animal pathogen, colonizing diverse ecological niches within its hosts. Predicting whether an isolate will infect a specific host and its subsequent clinical fate remains unknown. In this study, we investigated the S. aureus pangenome using a curated set of 356 strains, spanning a wide range of hosts, origins, and clinical display and antibiotic resistance profiles. We used genome-wide association study (GWAS) and random forest (RF) algorithms to discriminate strains based on their origins and clinical sources. Here, we show that the presence of sak and scn can discriminate strains based on their host specificity, while other genes such as mecA are often associated with virulent outcomes. Both GWAS and RF indicated the importance of intergenic regions (IGRs) and coding DNA sequence (CDS) but not sRNAs in forecasting an outcome. Additional transcriptomic analyses performed on the most prevalent clonal complex 8 (CC8) clonal types, in media mimicking nasal colonization or bacteremia, indicated three RNAs as potential RNA markers to forecast infection, followed by 30 others that could serve as infection severity predictors. Our report shows that genetic association and transcriptomics are complementary approaches that will be combined in a single analytical framework to improve our understanding of bacterial pathogenesis and ultimately identify potential predictive molecular markers. IMPORTANCE Predicting the outcome of bacterial colonization and infections, based on extensive genomic and transcriptomic data from a given pathogen, would be of substantial help for clinicians in treating and curing patients. In this report, genome-wide association studies and random forest algorithms have defined gene combinations that differentiate human from animal strains, colonization from diseases, and nonsevere from severe diseases, while it revealed the importance of IGRs and CDS, but not small RNAs (sRNAs), in anticipating an outcome. In addition, transcriptomic analyses performed on the most prevalent clonal types, in media mimicking either nasal colonization or bacteremia, revealed significant differences and therefore potent RNA markers. Overall, the use of both genomic and transcriptomic data in a single analytical framework can enhance our understanding of bacterial pathogenesis.

Project description:Mammographic density (MD) is an established predictor of risk of a first breast cancer, but the relationship of MD to contralateral breast cancer (CBC) risk is not clear, including the roles of age, mammogram timing, and change with treatment. Multivariable prediction models for CBC risk are needed and MD could contribute to these.We conducted a case-control study of MD and CBC risk in phase II of the WECARE study where cases had a CBC diagnosed ??2 years after first diagnosis at age <55 years and controls had unilateral breast cancer (UBC) with similar follow-up time. We retrieved film mammograms of the unaffected breast from two time points, prior to/at the time of the first diagnosis (253 CBC cases, 269 UBC controls) and ??6 months up to 48 months following the first diagnosis (333 CBC cases, 377 UBC controls). Mammograms were digitized and percent MD (%MD) was measured using the thresholding program Cumulus. Odds ratios (OR) and 95% confidence intervals (CI) for association between %MD and CBC, adjusted for age, treatment, and other factors related to CBC, were estimated using logistic regression. Linear regression was used to estimate the association between treatment modality and change in %MD in 467 women with mammograms at both time points.For %MD assessed following diagnosis, there was a statistically significant trend of increasing CBC with increasing %MD (p = 0.03). Lower density (<25%) was associated with reduced risk of CBC compared to 25 to <?50% density (OR 0.69, 95% CI 0.49, 0.98). Similar, but weaker, associations were noted for %MD measurements prior to/at diagnosis. The relationship appeared strongest in women aged <?45 years and non-existent in women aged 50 to 54 years. A decrease of ??10% in %MD between first and second mammogram was associated marginally with reduced risk of CBC (OR 0.63, 95% CI 0.40, 1.01) compared to change of <10%. Both tamoxifen and chemotherapy were associated with statistically significant 3% decreases in %MD (p <?0.01).Post-diagnosis measures of %MD may be useful to include in CBC risk prediction models with consideration of age at diagnosis. Chemotherapy is associated with reductions in %MD, similar to tamoxifen.

Dataset Information

Machine learning on genome-wide association studies to predict the risk of radiation-associated contralateral breast cancer in the WECARE Study.

Publications

Machine learning on genome-wide association studies to predict the risk of radiation-associated contralateral breast cancer in the WECARE Study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets