Prediction and Analysis of Skin Cancer Progression using Genomics Profiles of Patients

Abstract

The metastatic Skin Cutaneous Melanoma (SKCM) has been associated with diminished survival rates and high mortality rates worldwide. Thus, segregating metastatic melanoma from the primary tumors is crucial to employ an optimal therapeutic strategy for the prolonged survival of patients. The SKCM mRNA, miRNA and methylation data of TCGA is comprehensively analysed to recognize key genomic features that can segregate metastatic and primary tumors. Further, machine learning models have been developed using selected features to distinguish the same. The Support Vector Classification with Weight (SVC-W) model developed using the expression of 17 mRNAs achieved Area under the Receiver Operating Characteristic (AUROC) curve of 0.95 and an accuracy of 89.47% on an independent validation dataset. This study reveals the genes C7, MMP3, KRT14, LOC642587, CASP7, S100A7 and miRNAs hsa-mir-205 and hsa-mir-203b as the key genomic features that may substantially contribute to the oncogenesis of melanoma. Our study also proposes genes ESM1, NFATC3, C7orf4, CDK14, ZNF827, and ZSWIM7 as novel putative markers for cutaneous melanoma metastasis. The major prediction models and analysis modules to predict metastatic and primary tumor samples of SKCM are available from a webserver, CancerSPP (http://webs.iiitd.edu.in/raghava/cancerspp/).

Computing Skin Cutaneous Melanoma Outcome From the HLA-Alleles and Clinical Characteristics

“…The modification of surface molecules, lack of co-stimulatory molecules, production of immunosuppressive cytokines, and alterations in HLA molecules are some of the primary escape mechanisms used by tumor cells to evade the immune response (Garrido et al, 2010), which can directly distress the survival of an individual. Previous studies reveal that cutaneous melanoma is one of the most threatening and fatal form of skin cancer and scrutinized multi-omics signatures for the progression of malignancy (Li et al, 2015;Ossio et al, 2017;Bhalla et al, 2019). Further, in the past, it has been shown that if melanoma is detected at an early stage, the OS rate is 95%; but, once it is metastasized (lesion thickness > 4 mm); they are tough to cure, and the survival rate is reduced to less than 50% (Büttner et al, 1995;Bristow et al, 2010).…”

Identification of Platform-Independent Diagnostic Biomarker Panel for Hepatocellular Carcinoma Using Large-Scale Transcriptomics Data

“…It is implemented using in-house R scripts after the assignment of samples to the respective class, i.e., cancer or normal. These tests have been applied previously in different studies for the identification of DEGs (WELCH, 1947;Akaiwa et al, 1999;Carvalho and Irizarry, 2010;Aino et al, 2014;Schulze et al, 2015;Best et al, 2016;Bhasin et al, 2016;Bhalla et al, 2017;Cai et al, 2017;Bhalla et al, 2019;Cai et al, 2019;Kaur et al, 2019). Wilcoxon T-test is used for paired samples and Welch T-test is used for unpaired samples.…”

“…To reduce the number of genes from the selected set of signature, i.e., "the core genes of hepatocellular carcinoma," genes were ranked on training dataset (GSE25097) using a simple thresholdbased approach (Bhalla et al, 2017;Bhalla et al, 2019;Kaur et al, 2019). In the threshold-based approach, genes with a score above the threshold are assigned to cancer if it is found to be upregulated in cancer and otherwise normal; whereas sample is assigned to normal if the gene is downregulated in cancerous condition.…”

“…In the current study, both internal and external validation techniques were employed to evaluate the performance of models. First, the training dataset is used to develop prediction models and standard 10-fold cross-validation is used for performing internal validation, which is commonly employed in the literature (Burton et al, 2012;Bastani et al, 2013;Kourou et al, 2015;Bhalla et al, 2017;Jiang et al, 2018;Bhalla et al, 2019;Kaur et al, 2019). It is important to evaluate the realistic performance of the model on the external validation dataset, which should not be used for training and testing during model development.…”

Expression based biomarkers and models to classify early and late-stage samples of Papillary Thyroid Carcinoma

“…From the above analysis, it has been observed that the prediction models based on both protein-coding and non-coding transcripts gave higher performance as compared to protein-coding and cancer hallmark protein-coding transcripts alone. One of our recent studies has shown that the prediction model based on the SVC-L1 feature selection method achieved higher performance with the minimum number of features [50]. Hence, we performed feature selection using the SVC with L1 penalty (see methods).…”