Identifying statistically significant combinatorial markers for survival analysis

Refining the genetic risk of breast cancer with rare haplotypes and pattern mining

Hundreds of common variants have been found to confer small but significant differences in breast cancer risk, supporting the widely accepted polygenic model of inherited predisposition. Using a novel closed-pattern mining algorithm, we provide evidence that rare haplotypes may refine the association of breast cancer risk with common germline alleles. Our method, called Chromosome Overlap, consists in iteratively pairing chromosomes from affected individuals and looking for noncontiguous patterns of shared alleles. We applied Chromosome Overlap to haplotypes of genotyped SNPs from female breast cancer cases from the UK Biobank at four loci containing common breast cancer-risk SNPs. We found two rare (frequency <0.1%) haplotypes bearing a GWAS hit at 11q13 (hazard ratio = 4.21 and 16.7) which replicated in an independent, European ancestry population at P < 0.05, and another at 22q12 (frequency <0.2%, hazard ratio = 2.58) which expanded the risk pool to noncarriers of a GWAS hit. These results suggest that rare haplotypes (or mutations) may underlie the "synthetic association" of breast cancer risk with at least some common variants.

Identification of a prognostic signature in colorectal cancer using combinatorial algorithm-driven analysis

Colorectal carcinoma is one of the most common types of malignancy and a leading cause of cancer‐related death. Although clinicopathological parameters provide invaluable prognostic information, the accuracy of prognosis can be improved by using molecular biomarker signatures. Using a large dataset of immunohistochemistry‐based biomarkers (n = 66), this study has developed an effective methodology for identifying optimal biomarker combinations as a prognostic tool. Biomarkers were screened and assigned to related subsets before being analysed using an iterative algorithm customised for evaluating combinatorial interactions between biomarkers based on their combined statistical power. A signature consisting of six biomarkers was identified as the best combination in terms of prognostic power. The combination of biomarkers (STAT1, UCP1, p‐cofilin, LIMK2, FOXP3, and ICOS) was significantly associated with overall survival when computed as a linear variable (χ² = 53.183, p < 0.001) and as a cluster variable (χ² = 67.625, p < 0.001). This signature was also significantly independent of age, extramural vascular invasion, tumour stage, and lymph node metastasis (Wald = 32.898, p < 0.001). Assessment of the results in an external cohort showed that the signature was significantly associated with prognosis (χ² = 14.217, p = 0.007). This study developed and optimised an innovative discovery approach which could be adapted for the discovery of biomarkers and molecular interactions in a range of biological and clinical studies. Furthermore, this study identified a protein signature that can be utilised as an independent prognostic method and for potential therapeutic interventions.

Identification of Potential Early Diagnostic Biomarkers of Sepsis

Objective

The goal of this article was to identify potential biomarkers for early diagnosis of sepsis in order to improve their survival.

Methods

We analyzed differential gene expression between adult sepsis patients and controls in the GSE54514 dataset. Coexpression analysis was used to cluster coexpression modules, and enrichment analysis was performed on module genes. We also analyzed differential gene expression between neonatal sepsis patients and controls in the GSE25504 dataset, and we identified the subset of differentially expressed genes (DEGs) common to neonates and adults. All samples in the GSE54514 dataset were randomly divided into training and validation sets, and diagnostic signatures were constructed using least absolute shrink and selection operator (LASSO) regression. The key gene signature was screened for diagnostic value based on area under the receiver operating characteristic curve (AUC). STEM software identified dysregulated genes associated with sepsis-associated mortality. The ssGSEA method was used to quantify differences in immune cell infiltration between sepsis and control samples.

Results

A total of 6316 DEGs in GSE54514 were obtained spanning 10 modules. Module genes were mainly enriched in immune and metabolic responses. Screening 51 genes from among common genes based on AUC > 0.7 led to a LASSO model for the training set. We obtained a 25-gene signature, which we validated in the validation set and in the GSE25504 dataset. Among the signature genes, SLC2A6, C1ORF55, DUSP5 and RHOB were recognized as key genes (AUC > 0.75) in both the GSE54514 and GSE25504 datasets. SLC2A6 was identified by STEM as associated with sepsis-associated mortality and showed the strongest positive correlation with infiltration levels of Th1 cells.

Conclusion

In summary, our four key genes may have important implications for the early diagnosis of sepsis patients. In particular, SLC2A6 may be a critical biomarker for predicting survival in sepsis.