Statistical Issues in the Design and Analysis of nCounter Projects.
1Department of Biostatistics and Bioinformatics, Duke University, Durham, NC, USA. ; Biostatistics and Clinical Epidemiology Center, Samsung Medical Center, Seoul, Korea.
Cancer Informatics
|January 10, 2015
Summary
New NanoString nCounter technology offers targeted gene expression analysis. This paper discusses statistical methods for designing and analyzing nCounter projects, considering its unique counting-based approach compared to traditional microarrays.
Area of Science:
- Genomics
- Bioinformatics
- Statistical Genetics
Background:
- Traditional microarray platforms (e.g., Affymetrix, Illumina) offer genome-wide gene expression but have manufacturer-selected gene sets, limiting disease-specific analysis.
- NanoString's nCounter technology allows customer-selected gene panels (up to 800 genes) for targeted expression analysis, often using genes identified from prior high-throughput studies.
- nCounter employs a counting-based observation method, distinct from the continuous observations of traditional high-throughput platforms, necessitating tailored statistical approaches.
Purpose of the Study:
- To review and discuss statistical methodologies applicable to the design and analysis of gene expression studies utilizing NanoString's nCounter platform.
- To highlight the differences between nCounter and traditional microarray platforms that impact statistical method selection.
- To provide guidance on appropriate statistical techniques for nCounter-based research projects.
Main Methods:
- Comparative analysis of statistical requirements for traditional microarrays versus NanoString nCounter technology.
- Discussion of existing statistical methods and potential modifications for count-based gene expression data.
- Exploration of design considerations specific to targeted gene panels in nCounter experiments.
Main Results:
- Identification of statistical methods suitable for nCounter data analysis, accounting for its discrete nature.
- Highlighting the limitations of applying traditional microarray statistical methods directly to nCounter data.
- Emphasis on the importance of careful gene selection and appropriate statistical controls in nCounter study design.
Conclusions:
- Statistical methods for nCounter projects require adaptation due to the technology's unique characteristics, particularly its counting-based output.
- Careful consideration of experimental design and the selection of appropriate statistical tools are crucial for robust nCounter-based research.
- This work provides a foundation for researchers using nCounter technology to ensure valid and reliable gene expression analysis.
