TCGAnalyzeR: A web portal for visualization of pan-cancer molecular patient data

The Cancer Genome Atlas (TCGA) database (TCGA Network, 2013) is one of the largest and most commonly used public resources. It contains multidimensional molecular data of 11,000 cancer patients. We carried out an integrated meta-analysis of the single-nucleotide variations (SNVs), the copy number variations (CNVs), RNA-seq and clinical data of lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), liver hepatocellular carcinoma (LIHC) and breast invasive carcinoma (BRCA) patients from TCGA and PubChem BioAssay databases. In this project, we have deployed a web platform enabling statistical analysis of big data in 4 main categories providing the users to select the cancer type, data category, mutation type, risk group and cohort type. 1) In the category of "SNV" analysis, 5 types of plots was implemented; Oncoplot, OncoPrint, Lollipop Plot, Pyramid Plot and Enrichment Plots; Bar Plot, Cnet Plot, Heat Plot and Upset Plot. 2) In the category of "CNV" analysis, 3 types of plots was implemented; CNV Plot, OncoPrint and Enrichment Plots; Bar Plot, Cnet Plot, Heat Plot and Upset Plot. 3) In the category of "Transcriptome Profiling" analysis, 4 types of plots was provided: Volcano Plot, Box Plot and Enrichment Plots; Bar Plot, Cnet Plot, Heat Plot and Upset Plot. 4) In the category of "Clinical" analysis, data distribution plots and survival plots was drawn using the column names that the user dynamically choose from the clinical data. Each plot has its own filtration options and data tables. Therefore the user can redraw the plot by applying the filters. Users also can download the plot and related tables. The gene names given in the tables are highlighted to enable the user to select and add a gene to the "My genes" panel for filtering plots and copying the selected genes to the clipboard. In addition to the molecular data categories, the "DrugFinder" page returns the most active protein target name of the selected drugs in TCGA containing the most number of PubChem experiments with bioactivity less than 1 micromolar. The prototype of the web platform was developed using the R programming language and deployed on a Linux Shiny server.