Figure 3 | Scientific Reports

Figure 3

From: Homologous repair deficiency score for identifying breast cancers with defective DNA damage response

Figure 3

Identifying the genetic features that determine HRD tumors. (A) The frequency of somatic alterations in tumors among the top 10% of HRD scores (HRD tumors). Genes were sorted by their truncating mutation count within HRD tumor patient group. The bar graph presents the list of genes with truncated somatic alterations in more than three tumors. (B) Genetic alterations in HRD tumor population. Of the 92 HRD tumors, 9 cases with BRCA1 germline mutations and 2 cases with BRCA2 germline mutations exhibited somatic alterations in DDR genes, and among these tumors, TP53 alteration was observed in 9 cases. Among the tumors with germline mutations in BRCA1 or BRCA2, there were two cases with an additional POLQ germline alteration and one case with a BRIP1 germline mutation. Subtype colour codes are as follows. Purple: luminal A, orange: luminal B, pink: Her2, red: basal-like, green: normal-like. (C) Differences in DDR pathway-associated expression observed in HRD tumors. Among the 88 DDR genes (redundant genes excluded), we observed a significant expression difference in genes in the HR, BER, and FA pathways. Expression values are colour-coded for each gene, the lowest expression value being bright green, the middle being black, and the highest being bright red. Tumor flags: Patients with HRD tumors are labelled black, and patients with non-HRD tumors are labelled maroon. Fold change: HRD tumor expression mean was divided by that of non-HRD tumors, and then a log2 transformation was applied, followed by colour-coding. Deep red indicates a high positive log2 fold change, and deep blue indicates a high negative log2 fold change; p-value: Differences in expression among groups were compared using the rank-sum test. Genes that had a nominal p-value < 0.05 were labelled red, and the others white; Adjusted p-value : For all nominal p-values, multiple testing correction was performed by adjusting the p-values using the default p.adjust function in R. (D) BRCA1 plays a more dominant role than BRCA2 in promoting development into HRD tumors. When the expression levels of the 30 genes involved in the HR pathway were compared for each group (HRD tumors vs. non-HRD tumors; BRCA2 germline mutation vs. BRCA1 germline mutation; BRCA1 germline mutation non-HRD tumors vs. BRCA1 WT non-HRD tumors; BRCA2 germline mutation non-HRD tumors vs. BRCA2 WT non-HRD tumors), 23 genes showed an identical pattern to that in BRCA1 germline mutation HRD tumors, while 7 genes showed an identical pattern to that in BRCA2 germline mutation tumors, and 5 genes showed an expression pattern specific to HRD tumors, irrespective of BRCA germline mutation. (E) High BARD1 and BRIP1 expression in HRD tumors. Comparing the expression levels of the 5 genes that showed a difference between HRD and non-HRD tumors showed significantly high levels for BARD1 (p = 5.2e−11) and BRIP1 (p = 2e−05) in HRD tumors. For calculating the p-value, the Wilcoxon rank-sum test was used, and the central line in the box plot represents the median.

Back to article page