TMM

TMM is an acronym used in multiple domains, and its meaning depends on the field of application. The two most common references are in statistics/bioinformatics and in software engineering.

In bioinformatics, Trimmed Mean of M-values (TMM) is a normalization method for RNA-Seq count data. Introduced by Robinson and Oshlack in 2010, TMM aims to adjust for differences in library size and compositional bias between samples. It computes sample-specific scaling factors by comparing log-fold changes (M-values) and mean expression levels (A-values) across a reference set of genes, after trimming extreme values. The resulting normalization factors are applied to counts prior to differential expression analysis, improving comparability across libraries. TMM is widely implemented in Bioconductor packages such as edgeR and is a standard step before downstream analyses.

In software engineering, TMM can refer to the Test Maturity Model, and its extension, TMMi. These models provide a structured framework to assess and improve software testing processes. They define maturity levels that guide governance, planning, test design, execution, and measurement. Organizations use TMM/TMMi to benchmark testing capability, align testing with business objectives, and plan targeted process improvements. The level progression typically ranges from initial or managed to defined, measured, and optimizing in the TMMi variant.

Other uses of the acronym exist in various domains, but the intended meaning is usually evident from context. In summary, TMM most commonly denotes the RNA-Seq normalization method or a software testing maturity framework.