Table 1.
List of mechanistic pathway activity methods
| Method | Date | Code | Pathway modeled | Circuit definition | Scoring method | Activation / inhibition | Input | Result | Scope |
|---|---|---|---|---|---|---|---|---|---|
| Hipathia [40] | 2017 | Web application | KEGG | Receptor-to-effector circuits | Propagation algorithm | Yes | MA, RNA-seq | P-value per circuit | Multiple analyses |
| http://hipathia.babelomics.org | |||||||||
| Hipathia R code | |||||||||
| MinePath [52] | 2016 | Web application | KEGG | All possible circuits | Discretized gene expression (GE) values with logical operators | Yes | MA RNA-seq | P-value per circuit | Multiple analyses |
| http://minepath.org/ | |||||||||
| subSPIA [53] | 2015 | R code | KEGG | Minimal spanning trees (MST) | Differentially expressed (DE) genes used to define the MST | No | MA RNA-seq | P-value of DE per circuit | Comparison |
| PathiVar [44] | 2015 | Web application | KEGG | Receptor-to-effector circuits | Probabilistic model | Yes | MA, VCF | P-value per circuit | Multiple analyses |
| http://pathivar.babelomics.org | |||||||||
| Pathome [54] | 2014 | NA | KEGG | Receptor-to-effector linear circuits | Correspondence between pattern activation/inhibition and co-expression in adjacent gene pairs | Yes | MA RNA-seq | P-value per circuit | Comparison |
| Pathiways [41, 42] | 2013 | Web application http://pathiways.babelomics.org | KEGG | Receptor-to-effector circuits | Probabilistic model | Yes | MA | P-value per circuit | Multiple analyses |
| DEAP [55] | 2013 | Python code | KEGG | Receptor-to-effector linear circuits | Running sum of discretized DE | Yes | MA RNA-seq | Maximally differential expressed pathway | Comparison |
| CLIPPER [37] and GraphiteWeb [56] | 2013 | R package | KEGG; Reactome | All possible circuits | Weighted sum of GE | No | MA RNA-seq | Most relevant circuit per pathway | Comparison |
| ToPASeq R package | |||||||||
| Web application: | |||||||||
| http://graphiteweb.bio.unipd.it/ | |||||||||
| TEAK [57] | 2013 | Code @ Google (Windows and Mac) | KEGG | Receptor-to-effector circuits | Fits a Bayesian network for circuit and uses the BIC | No | MA | Ranked circuits | Comparison |
| PRS [58] | 2012 | ToPASeq R package | KEGG | Trees of associated DE genes | Topologically weighted sum of DE | No | MA RNA-seq | Ranked subpathways | Comparison |
| DEGraph [36] | 2012 | R package | KEGG; User defined pathways | All possible circuits | Multivariate two-sample tests of means of DE genes within a subgraph. | No | MA RNA-seq | P-value of DE per circuit | Comparison |
| ToPASeq R package | |||||||||
| Rivera et al. [59] | 2012 | NA | NetPath | All possible circuits | Weighted Z-score of genes within the subgraph | No | MA | P-value of most perturbed circuit | Comparison |
| Chen et al. [60] | 2011 | NA | KEGG | Receptor-to-effector circuits | Euclidian distance | No | MA | P-value per circuit | Comparison |
| PWEA [61] | 2010 | ToPASeq R package | User defined pathways | All possible circuits | Mutual influence among gene expression within the circuit | No | MA RNA-seq | P-value of DE per pathway | Comparison |
| TopologyGSA [38] | 2010 | ToPASeq R package | User defined pathways | All possible circuits | Comparison of covariance matrices of genes in the circuit | Yes | MA RNA-seq | P-value of DE per pathway | Comparison |
| DEGAS [62] | 2010 | Java (Windows) | KEGG | All possible circuits | Heuristic to find the largest dysregulated circuit | No | MA | One circuit per pathway | Comparison |
| TAPPA [39] | 2007 | ToPASeq R package | KEGG | All possible circuits | Scores of co-expression that explain the compared conditions | No | MA RNA-seq | P-value of DE per pathway | Multiple analyses |
The first column (Method) contains the name or acronym of the method, if exists, otherwise, we refer to it as the first author of the publication. The second column (Date) contains the publication date. The third column (Code) informs on the availability of the code to run the method. The fourth column (Pathway modeled) indicates the database used for pathway definition used by the method. The fifth column (Circuit definition) is the type of circuit used by the method. The sixth column (scoring method) summarizes how the circuit activity is inferred in the method. The seventh column (Activation/inhibition) denotes whether the scoring method uses the information of activation or inhibition nodes. The eight column (Input) indicates the data type that inputs the method (MA: expression microarray; RNA-seq: counts of RNA-seq experiments; VCF: mutation files). The ninth column (Result) describes the results provided by the method. And the tenth column (Scope) indicates the type of analyses the method permits, which can be either only conventional two conditions comparison or a wide range of analyses if the method first recodes gene expression into circuit activities.