A comparison of per sample global scaling and per gene normalization methods for differential expression analysis of RNA-seq data

Normalization is an essential step with considerable impact on high-throughput RNA sequencing (RNA-seq) data analysis. Although there are numerous methods for read count normalization, it remains a challenge to choose an optimal method due to multiple factors contributing to read count variability t...

Full description

Saved in:
Bibliographic Details
Published in:PloS one 2017-05, Vol.12 (5), p.e0176185-e0176185
Main Authors: Li, Xiaohong, Brock, Guy N, Rouchka, Eric C, Cooper, Nigel G F, Wu, Dongfeng, O'Toole, Timothy E, Gill, Ryan S, Eteleeb, Abdallah M, O'Brien, Liz, Rai, Shesh N
Format: Article
Language:English
Subjects:
Abundance
Alternative splicing
Area Under Curve
Aroma
Assembly
Bias
Bioinformatics
Biology and Life Sciences
Biomarkers
Brain
Breast
Breast cancer
Breast Neoplasms - metabolism
Cardiology
Color
Comparative studies
Computer programs
Computer Simulation
Construction
Data analysis
Data processing
Datasets as Topic
Diagnosis
DNA microarrays
Engineering
Epidemiology
Error analysis
Exons
Experimental design
Factor analysis
Fragmentation
Gene expression
Gene Expression Profiling - methods
Gene mapping
Genomes
High-Throughput Nucleotide Sequencing - methods
Humans
Informatics
Instrumentation
Isoforms
Luteinizing hormone
Mapping
Mathematical analysis
Mathematical models
Medicine and Health Sciences
Methods
Microarray Analysis - methods
miRNA
Models, Statistical
Nervous system
Packages
Pattern recognition
Physical Sciences
Quality control
Radiology
Regression analysis
Research and Analysis Methods
Ribonucleic acid
RNA
RNA editing
RNA sequencing
ROC Curve
Samples
Scaling
Sequence Analysis, RNA - methods
Software
Statistical analysis
Statistical methods
Tissues
Transcription
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Normalization is an essential step with considerable impact on high-throughput RNA sequencing (RNA-seq) data analysis. Although there are numerous methods for read count normalization, it remains a challenge to choose an optimal method due to multiple factors contributing to read count variability that affects the overall sensitivity and specificity. In order to properly determine the most appropriate normalization methods, it is critical to compare the performance and shortcomings of a representative set of normalization routines based on different dataset characteristics. Therefore, we set out to evaluate the performance of the commonly used methods (DESeq, TMM-edgeR, FPKM-CuffDiff, TC, Med UQ and FQ) and two new methods we propose: Med-pgQ2 and UQ-pgQ2 (per-gene normalization after per-sample median or upper-quartile global scaling). Our per-gene normalization approach allows for comparisons between conditions based on similar count levels. Using the benchmark Microarray Quality Control Project (MAQC) and simulated datasets, we performed differential gene expression analysis to evaluate these methods. When evaluating MAQC2 with two replicates, we observed that Med-pgQ2 and UQ-pgQ2 achieved a slightly higher area under the Receiver Operating Characteristic Curve (AUC), a specificity rate > 85%, the detection power > 92% and an actual false discovery rate (FDR) under 0.06 given the nominal FDR (≤0.05). Although the top commonly used methods (DESeq and TMM-edgeR) yield a higher power (>93%) for MAQC2 data, they trade off with a reduced specificity (
ISSN:1932-6203
1932-6203
DOI:10.1371/journal.pone.0176185