Table 1. List of methods for Pathway Analysis.
| Method | Date | Code | Pathway modelled | Entity modelled | Input | Output | Comparison | Loops |
|---|---|---|---|---|---|---|---|---|
| MinePath[52] | 2015 | Web application http://minepath.org/ |
KEGG pathways | Subpath identification | MA | p-value per pathway p-value per subpathway binary value per sample graphical visualization |
Two conditions | NA |
| Qin et al.[53] | 2015 | NAb | 12 cancer-related KEGG pathways | signal quantification | Mutations CNVs Cancer drugs |
Pathway activity | Personalized | yes |
| subSPIA[13] | 2015 | R code | KEGG pathways | signal quantification | MA RNAseq (via SPIA in ToPASeq) |
p-value of DE per subpathway p-value of PF per subpathway global p-value (DE+PF) |
Two conditions | no |
| Pathome[54] | 2014 | NA | KEGG pathways | signal quantification | MA RNAseq |
p-value per subpathway | Two conditions | NA |
| Pepe et al.[55] | 2014 | R code | KEGG pathways | subpath identification | MA | p-value per subpathway | Two conditions | NA |
| ToPaSeq[18] | 2014 | R package | graphite gene-gene networks user's pathways |
integrates other methods: TopologyGSA DEGraph Clipper SPIA TAPPA PRS PWEA |
MA RNAseq |
Depends on the method | Two conditions | Depends on the method |
| DEAP[12] | 2013 | python code | user defined pathway structure | signal quantification | MA RNAseq |
Score and p-value per pathway subgraph with the maximum absolute score |
Two conditions | yes |
| CliPPER[5] | 2013 | R package ToPASeq R package |
graphite gene-gene networks cliques user's pathways (via ToPASeq) |
subpath identification | MA RNAseq |
p-value at pathway level Most affected subgraph per pathway Gene-level statistics for DE of genes |
Two conditions | no |
| GraphiteWeb[56] | 2013 | Web application: http://graphiteweb.bio.unipd.it/Rpackage |
KEGG pathways Reactome pathways |
integrates other methods: Hypergeometric test Global Test GSEA SPIA CliPPER |
MA RNAseq |
Significant pathways Visualization of the pathways with nodes coloured according to their contribution to the analysis |
Two conditions | no |
| TEAK[57] | 2013 | Code @ Google (Windows and Mac) | KEGG pathways | metabolism-orientedsubpathway identification | MA | Ranked subpathways | Two conditions | no |
| PRS[16] | 2012 | ToPASeq R package | graphite gene-gene networks (ToPASeq) user's pathways (via ToPASeq) |
pathway identification | MA RNAseq |
p-value per pathway gene-level statistics for DE of genes |
Two conditions | yes |
| DEGraph[6] | 2012 | R packageToPASeq R package | subgraphs of a large graph (branch-and-bound-like approach) graphite gene-gene networks (ToPASeq) user's pathways (via ToPASeq) |
subpath identification | MA RNAseq |
p-value of DE per subpathway p-value per pathway Gene-level statistics for DE of genes |
Two conditions | no |
| Rivera et al.[58] | 2012 | NA | NetPathpathways | subpath identification | MA | p-value of most perturbed subpathway | Two conditions | NA |
| Chen et al.[59] | 2011 | NA | KEGG pathways | subpath identification | MA | p-value per subpathway p-value of key genes |
Two conditions | NA |
| PWEA[17] | 2010 | ToPASeq R package | Complete pathways (KEGG) graphite gene-gene networks (ToPASeq) user's pathways (via ToPASeq) |
pathway identification | MA RNAseq |
p-value of DE per pathway Gene-level statistics for DE of genes |
Two conditions | no |
| TopologyGSA[14] | 2010 | ToPASeq R package | Complete pathways (KEGG) Cliques graphite gene-gene networks (ToPASeq) user's pathways (via ToPASeq) |
subpath identification | MA RNAseq |
p-value of DE per pathway Gene-level statistics for DE of genes |
Two conditions | no |
| DEGAS[60] | 2010 | Java (Windows) | KEGG pathways PPIs network |
novel subpath identification | MA | A subpathway per pathway | Two conditions | NA |
| TAPPA[15] | 2007 | ToPASeq R package | graphite gene-gene networks (ToPASeq) user's pathways (via ToPASeq) |
pathway identification | MA RNASeq |
p-value of DE per pathway Gene-level statistics for DE of genes |
Two conditions | no |
The first column (Method) contains the name or acronym of the method, if exists, otherwise, we refer to it as the fires 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 modelled) indicates the pathway definition used in the method. The fifth column (Entity modelled) is the entity, within the pathway, used in the method (“subpath identification” methods obtain candidate sub-pathways usually by differential expression of its constituent genes, “signal quantification” methods provide, in addition, a quantification of the activation status of the sub-pathway). The sixth column (input) indicates the data type that inputs the method (MA: Expression Microarray; CNV: copy number variation; NA: not available). The seventh column (output) describes the results provided by the method. Some provide only a score (p-value, DE: differential expression matrix; PF: perturbation factor) for the whole pathway and other also provide scores for sub-pathways, that can be defined within the pathways in many different ways. The eight column (Comparison) indicates the type of comparison the method can deal with. It can be either a conventional two conditions (typically case/control) comparison or it can allow obtaining personalized results per individual. And the ninth column (Loops) indicates whether the method can handle loop structures in the topology of the sub-pathway analysed or not.