Identification of a gene expression signature associated with breast cancer survival and risk that improves clinical genomic platforms.

Motivation: Modern genomic technologies allow us to perform genome-wide analysis to find gene markers associated with the risk and survival in cancer patients. Accurate risk prediction and patient stratification based on robust gene signatures is a key path forward in personalized treatment and precision medicine. Several authors have proposed the identification of gene signatures to assign risk in patients with breast cancer (BRCA), and some of these signatures have been implemented within commercial platforms in the clinic, such as Oncotype and Prosigna. However, these platforms are black boxes in which the influence of selected genes as survival markers is unclear and where the risk scores provided cannot be clearly related to the standard clinicopathological tumor markers obtained by immunohistochemistry (IHC), which guide clinical and therapeutic decisions in breast cancer. Results: Here, we present a framework to discover a robust list of gene expression markers associated with survival that can be biologically interpreted in terms of the three main biomolecular factors (IHC clinical markers: ER, PR and HER2) that define clinical outcome in BRCA. To test and ensure the reproducibility of the results, we compiled and analyzed two independent datasets with a large number of tumor samples (1024 and 879) that include full genome-wide expression profiles and survival data. Using these two cohorts, we obtained a robust subset of gene survival markers that correlate well with the major IHC clinical markers used in breast cancer. The geneset of survival markers that we identify (which includes 34 genes) significantly improves the risk prediction provided by the genesets included in the commercial platforms: Oncotype (16 genes) and Prosigna (50 genes, i.e. PAM50). Furthermore, some of the genes identified have recently been proposed in the literature as new prognostic markers and may deserve more attention in current clinical trials to improve breast cancer risk prediction. Availability and implementation: All data integrated and analyzed in this research will be available on GitHub (https://github.com/jdelasrivas-lab/breastcancersurvsign), including the R scripts and protocols used for the analyses. Supplementary information: Supplementary data are available at Bioinformatics Advances online.

Survival marker genes of colorectal cancer derived from consistent transcriptomic profiling.

BACKGROUND: Identification of biomarkers associated with the prognosis of different cancer subtypes is critical to achieve better therapeutic assistance. In colorectal cancer (CRC) the discovery of stable and consistent survival markers remains a challenge due to the high heterogeneity of this class of tumors. In this work, we identified a new set of gene markers for CRC associated to prognosis and risk using a large unified cohort of patients with transcriptomic profiles and survival information. RESULTS: We built an integrated dataset with 1273 human colorectal samples, which provides a homogeneous robust framework to analyse genome-wide expression and survival data. Using this dataset we identified two sets of genes that are candidate prognostic markers for CRC in stages III and IV, showing either up-regulation correlated with poor prognosis or up-regulation correlated with good prognosis. The top 10 up-regulated genes found as survival markers of poor prognosis (i.e. low survival) were: DCBLD2, PTPN14, LAMP5, TM4SF1, NPR3, LEMD1, LCA5, CSGALNACT2, SLC2A3 and GADD45B. The stability and robustness of the gene survival markers was assessed by cross-validation, and the best-ranked genes were also validated with two external independent cohorts: one of microarrays with 482 samples; another of RNA-seq with 269 samples. Up-regulation of the top genes was also proved in a comparison with normal colorectal tissue samples. Finally, the set of top 100 genes that showed overexpression correlated with low survival was used to build a CRC risk predictor applying a multivariate Cox proportional hazards regression analysis. This risk predictor yielded an optimal separation of the individual patients of the cohort according to their survival, with a p-value of 8.25e-14 and Hazard Ratio 2.14 (95% CI: 1.75-2.61). CONCLUSIONS: The results presented in this work provide a solid rationale for the prognostic utility of a new set of genes in CRC, demonstrating their potential to predict colorectal tumor progression and evolution towards poor survival stages. Our study does not provide a fixed gene signature for prognosis and risk prediction, but instead proposes a robust set of genes ranked according to their predictive power that can be selected for additional tests with other CRC clinical cohorts.

κ-NN for the classification of human cancer samples using the gene expression profiles.

The [Formula: see text]-Nearest Neighbor (k-NN) classifier has been applied to the identification of cancer samples using the gene expression profiles with encouraging results. However, the performance of [Formula: see text]-NN depends strongly on the distance considered to evaluate the sample proximities. Besides, the choice of a good dissimilarity is a difficult task and depends on the problem at hand. In this chapter, we introduce a method to learn the metric from the data to improve the [Formula: see text]-NN classifier. To this aim, we consider a regularized version of the kernel alignment algorithm that incorporates a term that penalizes the complexity of the family of distances avoiding overfitting. The error function is optimized using a semidefinite programming approach (SDP). The method proposed has been applied to the challenging problem of cancer identification using the gene expression profiles. Kernel alignment [Formula: see text]-NN outperforms other metric learning strategies and improves the classical [Formula: see text]-NN algorithm.

Combining dissimilarities in a Hyper Reproducing Kernel Hilbert Space for complex human cancer prediction.

DNA microarrays provide rich profiles that are used in cancer prediction considering the gene expression levels across a collection of related samples. Support Vector Machines (SVM) have been applied to the classification of cancer samples with encouraging results. However, they rely on Euclidean distances that fail to reflect accurately the proximities among sample profiles. Then, non-Euclidean dissimilarities provide additional information that should be considered to reduce the misclassification errors. In this paper, we incorporate in the nu-SVM algorithm a linear combination of non-Euclidean dissimilarities. The weights of the combination are learnt in a (Hyper Reproducing Kernel Hilbert Space) HRKHS using a Semidefinite Programming algorithm. This approach allows us to incorporate a smoothing term that penalizes the complexity of the family of distances and avoids overfitting. The experimental results suggest that the method proposed helps to reduce the misclassification errors in several human cancer problems.