Table 1.
GP2S (Stegle at al. 2010) |
TCAP (Kiddle et al. 2010) |
|
---|---|---|
Advantages | Effectively acts as a filter to remove genes/metabolites which do not change in expression /concentration over time Computationally fast Developed to be robust to outliers Uses data from all biological replicates (not mean) to produce models Synchronized observation times are not required Can be extended to elucidate the time at which differential expression/metabolite concentration differences begin to occur |
Information–rich similarity measure and clustering algorithm finds gene expression / metabolite concentrations which follow the same time series or incorporates temporal changes: a) Delays—a similar time series profile but with a lag b) Transient correlations—similar time series for some time points, not all c) Inversions—same time series profile but inverted d) Combination of the above Robust, with little user input Output simple to interpret |
Disadvantages | Only compares two treatment groups at a time | Mean time series of replicates required Whilst faster than other transcriptomics clustering algorithms, it is computationally intensive |