Skip to main content

Table 2 Description of various software packages for conducting differential abundance tests

From: The importance of study design for detecting differentially abundant features in high-throughput experiments

Name

Statistical testing

Normalization approach

Target application areas

edgeR [12]

Negative Binomial Model, Conditional Maximum Likelihood to estimate parameters, Exact Test

TMM, UQN

SAGE [1], MPSS [29], PMAGE [30], miRAGE [31] and SACO [32] among others

DESeq [6]

Negative Binomial Model, local regression to estimate parameters, Exact Test

Normalization by median

RNA-seq [3], HITS-CLIP [9] and ChIP-seq [33] among others

baySeq [15]

Negative Binomial Model, empirical Bayes to estimate parameters

RPM, TMM

DNA-seq, RNA-seq [3] and SAGE [1] among others

NOISeq [16]

Non-parametric approach

RPM, RPKM, UQN

ChIP-seq [33] and RNA-seq [3] among others

Cuffdiff [14]

t-test

RPM, RPKM, UQN

RNA-seq [3]

Metastats [11]

Non-parametric t-test, Fisher’s Exact Test for small counts

RPM

Metagenomics

  1. RPKM: Normalization by read count and gene length (RNA-seq) [3], RPM: Normalization by read count (RNA-seq), TMM: Trimmed Mean of M values [34], UQN: Upper Quartile Normalization [20].