Integrative Statistical Analysis Pipeline for RNA-Seq and NanoString with Application to Gene Expression Data of Cancer Patients

• Activity Number: 368 - SPEED: Statistics for Biopharmaceutical Studies
• Type: Contributed
• Date/Time: Tuesday, July 31, 2018 : 11:35 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Jeea Choi* and Catarina D. Campbell and Xiaoshan Wang and He Wei and Robinson Douglas and Stephane Wong and Bin Fu and Rebecca Leary and Kavitha Venkatesan and Ying A Wang
• Companies: Novartis Pharmaceuticals and Novartis Institutes for BioMedical Research and Novartis Pharmaceuticals and Novartis Pharmaceuticals and Novartis Pharmaceuticals and Novartis Pharmaceuticals and Novartis Pharmaceuticals and Novartis Institutes for BioMedical Research and Novartis Institutes for BioMedical Research and Novartis Pharmaceuticals
• Keywords: Biomarker; NanoString; RNA-sequencing; non-small cell lung cancer; melanoma; tumor microenvironment

Abstract

Gene expression profiling enables a wide variety of basic research, translational medicine, and in vitro diagnostics applications. Over the past several years, RNA-sequencing has become the standard technique for transcriptome-wide gene expression profiling. Other techniques, such as NanoString's nCounter technology allow for targeted gene expression assessment on a subset of genes. Many variations of statistical methods for RNA-seq and NanoString have been published, which brings challenges to new users. With the goal to provide a comprehensible statistical analysis pipeline, we implemented major analyses steps in current standards for both RNA-seq and NanoString analysis into a R package called "RNAexpressionToolbox". We applied the pipeline to both RNA-seq and NanoString expression data from commercial samples of 38 melanoma cancer patients and 29 non-small cell lung cancer patients, independently. We compared the RNA-seq and NanoString platforms from various perspectives including number of genes with detectable expression and concordance of gene expression measurements. In addition, we explored the tumor microenvironment in melanoma and NSCLS based on immune-related genes.