New discovery of breast cancer! Identify 32 new susceptibility gene constellations.

Today, I'd like to share with you an article published in Nature Genetics (if: 25.455) in June: genome wide association study identities 32 new breast cancer susceptibility loci from overall and subtype specific We conducted a genome-wide association study using standard and new methods to explain potential tumor heterogeneity based on the status and tumor grade of estrogen receptor, progesterone receptor and human epidermal growth factor receptor 2, and identified 32 new susceptibility loci.genome wide association study identifications 32 new breast cancer susceptibility loci from overall and subtype specific analyses were used to identify 32 new breast cancer susceptibility loci (locus, loci): the position of genes on chromosomes.at the molecular level, it is a DNA sequence with genetic effects.vividly speaking, a pair of chromosomes can be imagined as two parallel lines. A given position on a chromosome is like a point or segment corresponding to two parallel lines, which is called a locus.a locus can be a gene, a part of a gene, or a DNA sequence with some regulatory effect.a locus is different from a site, which is a mutation site within a cis trans, and can be as small as a nucleotide pair.a locus is a fixed site on a chromosome. Alleles encode the same DNA at the same locus.alleles at some loci have significant individual differences, so they can be used to identify a person like a fingerprint.genome wide association Studies (GWAS) refers to the multi center, large sample and repeatedly verified association study between genes and diseases at the whole genome level. It is a research method to find the genetic factors related to complex diseases by genotyping large-scale population DNA samples with high-density genetic markers (such as SNP or CNV), so as to comprehensively reveal the occurrence, development and treatment of diseases It's a genetic gene that we're not. The susceptibility variation of breast cancer (BRCA) often shows heterogeneity in tumor subtypes.in order to identify the new loci, a genome-wide association study was conducted in 133384 breast cancer cases and 113789 controls, as well as 18908 BRCA1 mutation carriers of European origin (9414 breast cancer patients) Tumor grade to explain the potential tumor heterogeneity.thirty two new susceptibility loci were identified, 15 of which were associated with at least one tumor feature.5 loci showed opposite correlation between luminal subtype and non luminal subtype.computer simulation analysis showed that these five loci contained cell specific enhancers, which were different from normal luminal and basal breast cells.the genetic correlation among the five Intransic like subtypes ranged from 0.35 to 0.80.all known susceptibility loci were explained by genome-wide microarray heritability. Luminal A-like and triple negative accounted for 54.2% and 37.6% respectively.the odds ratio (or) of polygenic risk scores (or), which is calculated by the genotype effect value of GWAS statistical data, includes 330 variants, luminal The odds ratios (including 330 variations) of the highest 1% quantile and the median of A-like and triple negative were 5.63 and 3.02, respectively.these findings improve the understanding of the genetic susceptibility of breast cancer subtypes and will provide information for the development of subtype specific polygenic risk scores. background according to the largest genome-wide association study (GWAS) of the breast cancer Association Alliance (BCAC), more than 170 independent breast cancer susceptibility variants have been identified. many of these variations show a correlation between different tumor subtypes, especially in estrogen receptor positive and estrogen receptor negative or triple negative diseases. however, previous GWAS did not consider the high correlation and classification of multiple tumor markers, such as estrogen receptor, progesterone receptor and human epidermal growth factor receptor 2 (HER2), to identify the specific source of etiological heterogeneity. in this work, we used standard analysis and a new two-stage multivariate regression (TWS) method to study GWAS of breast cancer. This method effectively characterizes the heterogeneity of etiology, and takes into account the correlation of tumor markers and missing data. 3. Data 1. Overall breast cancer samples: 82 BCAC studies of women of European origin from 20 countries. genotyping data were obtained from two Illumina genome-wide custom arrays: icogs and oncoarray (supplementary Table 1). the total sample size analyzed (including icogs, oncoarray and other GWAS data) was 133384 samples and 113789 controls. GWAS data: two stage logistic regression analysis was performed using 106278 invasive samples and 91477 controls. Cimba data: there were 15566 BRCA1 mutation carriers over 18 years old, and 7784 of them had breast cancer (supplementary Table 3). the icogs genotype data of 3342 BRCA1 mutation carriers (1630 breast cancer patients) in 54 studies were also obtained through cimba. The odds ratio (ORS) and 95% confidence interval (CI) were estimated by standard logistic regression, and adjusted according to the country and principal component. the icogs and oncoarray data were evaluated respectively, and fixed effects, one of the basic concepts of experimental design, were used. in the experiment, if the level of the factor has been determined, the horizontal effect of the factor is taken as a fixed parameter, which is called the fixed effect). 2. In order to identify evidence of heterogeneous breast cancer susceptibility variations, a new score test based on two-level multiple regression was used. The model allows flexible and concise association models in the presence of potential heterogeneity in estrogen receptor, progesterone receptor, HER2 and / or grade. the model implements an effective expectation maximization algorithm to deal with the missing tumor feature data. these analyses were limited to BCAC controls and invasive cases. In this study, an additional two-level model was used to estimate the case-control ors and 95% confidence interval between the variant and the intrinsic like subtypes defined by estrogen receptor, progesterone receptor, HER2 and grade. (1) luminal A-like (2) luminal B / HER2 negative like (3) luminal B-like (4) HER2 enriched like (5) triple negative or basic like. the data of icogs and oncoarray were analyzed, and the principal component and age were adjusted, and the results were analyzed by using fixed effect model. a leave one out sensitivity analysis was used to assess the country's impact. The risk ratio of each allele was estimated within the framework of a retrospective cohort analysis among BRCA1 mutation carriers who were prone to develop triple negative disease. assume that the estimated ors of BCAC triple negative cases and the estimated risk ratio of cimba BRCA1 carriers are approximately the same potential relative risks, and the fixed effects meta-analysis is used to consolidate these results. in all new variants, a two-level multivariate model was used to test the association heterogeneity, global and tumor specific between different subtypes. Candidate causal variants (CCVs) are defined as the variation within ± 500kb of lead variation in each new region, and the p value is within 100 times of the lead variation. Results in general, 32 new independent susceptibility loci were identified; 5.0 × 10 − 8 (Figure 1): 22 variants were identified by standard logistic regression, 16 were identified by two-level multivariate model (8 of which were not detected by standard logistic regression), and 3 variables were identified in cimba / BCAC triple negative meta-analysis (rs78378222 was also detected by BCAC's two-level multivariate model). Figure 1. Meta analysis of the population, subtypes, BCAC triple negative and cimba BRCA1 carriers of all independent genome-wide significant breast cancer susceptibility variants showed heterogeneity in 15 out of 32 (Fig. 2). estrogen receptor (7 variants) and grade (7 variants) were the most common causes of observed heterogeneity, followed by HER2 (4 variants) and progesterone receptor (2 variants). figure 2.the thermogram and cluster p value of marker specific heterogeneity test of 32 breast cancer susceptibility loci showed the opposite correlation with luminal subtype and non luminal subtype (Fig. 3). the four variants were associated with luminal A-like and triple negative subtypes in the opposite direction. rs78378222 was associated with estrogen receptor and HER2; rs206435 was associated with estrogen receptor and grade; rs141526427 and rs6065254 were associated with estrogen receptor only. rs7924772 showed an opposite case-control association between HER2 negative and HER2 positive subtypes, consistent with these findings, rs7924772 was only associated with HER2 (Fig. 3). Figure 3. Susceptibility variations with opposite associations between different subtypes. Next, candidate causal variants (CCVs) were defined for each new site, and the relationship between CCVs in primary breast cells and previously annotated enhancers was studied. combined with h3k4me1 and h3k27ac histone modified chromatin immunoprecipitation sequencing (chip SEQ) signals, putative enhancers in basal cells, luminal progenitor cells and mature luminal cells were identified as off, primed and active. switch enhancers are defined as those that exhibit different characteristics between cell types. Among the five loci with reverse association, each locus has at least one CCV overlapping with one switch enhancer (Fig. 4). these results suggest that some variants may modulate enhancer activity in a cell type specific manner, thereby differentially affecting the risk of tumor subtypes. Figure 4. In five loci with opposite association directions between subtypes of primary breast cancer, CCVs overlapped with enhancer status used inquire to overlap CCVs with functional annotation data from a public database to identify potential target genes. 179 independent target genes were predicted for 26 of the 32 independent signals. it has been reported that rs78378222 is related to the level of TP53 mRNA in blood and adipose tissue, but it has not been repeated in breast tissue in this work. however, this study found that rs78378222 overlapped with a cell type specific regulatory element in breast basal epithelial cells, suggesting that enhancer function is another potential TP53 transcription control mechanism. 23 target genes in 14 regions were predicted with high confidence, and 22 target genes in 13 regions were predicted as distal regulation. inquiry previously predicted four target genes (polr3c, rnf115, SOX4 and Tbx3 (Tbx3 is a known driver gene for somatic breast cancer), and