EMTscore - Calculate EMT Scores Based on Omics Data
Epithelial-mesenchymal transition (EMT) is an important
form of cellular plasticity that is fully or partially
activated in several biological scenarios including development
and disease progression. EMT involves altered expression of
hundreds of protein-coding and non-protein-coding genes. Recent
studies showed the prevalence of partial EMT in multiple
processes such as various cancers and organ fibrosis, which
necessitates rigorous quantification of the degree of EMT.
While traditional gene set scoring methods such as gene set
variation analysis have been used to generate EMT scores from
omics data, multiple EMT scoring algorithms and EMT gene sets
have been used by different groups without standardization.
Furthermore, comparisons of EMT scores computed from different
methods and/or different EMT gene sets are generally difficult
due to both the context dependent nature of EMT and the lack of
tools that comprehensively integrate varying components for EMT
scoring. To address this problem, EMTscore enables users to
select scoring methods from a list of previously used
algorithms and EMT gene sets from a list of gene sets produced
from different experiments. Several visualization methods are
provided for making publication-quality plots of EMT scores
from omics data. The package also implements a
principal-component-analysis based method for scoring divergent
EMT processes from a single dataset. Overall, EMTscore provides
an integrated solution for assessing the degree and complexity
of EMT from omics data, and paves the way for standardizing the
comparison of EMT programs across multiple contexts.
