Table 3. Bipartite graphs.
Index | Data type 1 | Data type 2 | Statistical method/
association type |
Multiple
testing correction |
Cutoff
(corrected P-value) |
Implementation | Description and Biological implication | Reference(s) |
---|---|---|---|---|---|---|---|---|
1 | PROBETYPE | SYMBOL | mapping | N/A | N/A | getSYMBOL function in lumi package | Reference mapping of all probes on the
chip to genes (platform-specific) |
14 |
2 | PROBE | SYMBOL | mapping | N/A | N/A | topTable function in limma | Mapping of experiment-specific probes to
genes |
18 |
3 | PROBE | wgcna | mapping | N/A | N/A | WGCNA | Indicates membership of specific probes
in enumerated WGCNA modules |
14 |
4 X | PROBE | PROBE | Connectivity (weight
in topological overlap matrix) |
N/A | Maximum
of 2000 per module |
Custom function: at least the top 10 % of
edges in the module network, capped at 2000 |
Subset of probes in WGCNA modules with
high connectivity |
14 |
5 | SYMBOL | reactomePW | mapping | N/A | N/A | lumi, ReactomePA (uses reactome.
db) |
Mapping of gene symbols to reactome
pathways in the current dataset |
14, 22 |
6 | SYMBOL | PalWangPW | mapping | N/A | N/A | WGCNA (uses PWLists included with
WGCNA) |
Mapping of gene symbols to a manually
curated list of pathways in the current dataset |
14 |
7 | SYMBOL | ImmunePW | mapping | N/A | N/A | WGCNA (uses ImmunePathwayLists
included with WGCNA) |
Mapping of gene symbols to a manually
curated list of immune pathways in the current dataset |
14 |
8 | SYMBOL | cellEx | mapping | N/A | N/A | WGCNA (uses BloodLists included
with WGCNA) |
Mapping of gene symbols to a manually
curated list of cell-type specific genes in the current dataset |
14 |
9 | SYMBOL | cellEx | mapping | N/A | N/A | CellMix (HaemAtlas dataset) | Mapping of gene symbols to the Watkins,
et al. list of cell-type specific genes in the current dataset |
19, Supp.
Ref 30 |
10 | SYMBOL | cellEx | mapping | N/A | N/A | CellMix (Abbas dataset) | Mapping of gene symbols to the Abbas,
et al. list of cell-type specific genes in the current dataset |
19, Supp.
Ref 11 |
11 | wgcna | baylor
(Chaussabel modules) |
Hypergeometric test | BH | <0.05 | WGCNA (UserListEnrichment
function with Chaussabel module genes as input) |
Enrichment of WGCNA modules for
Chaussabel modules, allowing functional comparison of the two approaches |
14 |
12 | baylor
(Chaussabel modules) |
reactomePW | Hypergeometric test | BH | <0.05 * | WGCNA (UserListEnrichment
function with Reactome genes as input) |
Enrichment of Chaussabel modules for
Reactome pathways, allowing functional annotation of the module |
14, 22 |
13 | baylor
(Chaussabel modules) |
PalWangPW | Hypergeometric test | BH | <0.05 ** | WGCNA (UserListEnrichment
function with a custom list of pathway genes as input) |
Enrichment of Chaussabel modules for a
manually curated list of pathways, allowing functional annotation of the module |
14 |
14 | baylor
(Chaussabel modules) |
ImmunePW | Hypergeometric test | BH | <0.05 *** | WGCNA (UserListEnrichment
function with a custom list of immune pathway genes as input) |
Enrichment of Chaussabel modules for a
manually curated list of immune pathways, allowing functional annotation of the module |
14 |
15 | wgcna | reactomePW | Hypergeometric test | BH | <0.05 * | WGCNA (UserListEnrichment
function with Reactome genes as input) |
Enrichment of WGCNA modules for
Reactome pathways, allowing functional annotation of the module |
14, 22 |
16 | wgcna | PalWangPW | Hypergeometric test | BH | <0.05 ** | WGCNA (UserListEnrichment
function with a custom list of pathway genes as input) |
Enrichment of WGCNA modules for a
manually curated list of pathways, allowing functional annotation of the module |
14 |
17 | wgcna | ImmunePW | Hypergeometric test | BH | <0.05 *** | WGCNA (UserListEnrichment
function with a custom list of immune pathway genes as input) |
Enrichment of WGCNA modules for a
manually curated list of immune pathways, allowing functional annotation of the module |
14 |
18 | baylor
(Chaussabel modules) |
cellEx | Hypergeometric test | BH | <0.05 | WGCNA (UserListEnrichment
function, uses BloodLists included with WGCNA) |
Enrichment of Chaussabel modules for a
manually curated list of cell-type specific genes, allowing interpretation of cellular context of the module |
14 |
19 | baylor
(Chaussabel modules) |
cellEx | Hypergeometric test | BH | <0.05 | WGCNA (UserListEnrichment
function, uses HaemAtlas list included with Cellmix) |
Enrichment of Chaussabel modules for a
manually curated list of cell-type specific genes, allowing interpretation of cellular context of the module |
14,
19, Supp.
Ref 30 |
20 | baylor
(Chaussabel modules) |
cellEx | Hypergeometric test | BH | <0.05 | WGCNA (UserListEnrichment
function, uses Abbas list included with Cellmix) |
Enrichment of Chaussabel modules for a
manually curated list of cell-type specific genes, allowing interpretation of cellular context of the module |
14,
19, Supp.
Ref 11 |
21 | wgcna | cellEx | Hypergeometric test | BH | <0.05 | WGCNA (UserListEnrichment
function, uses BloodLists included with WGCNA) |
Enrichment of WGCNA modules for a
manually curated list of cell-type specific genes, allowing interpretation of cellular context of the module |
14 |
22 | wgcna | cellEx | Hypergeometric test | BH | <0.05 | WGCNA (UserListEnrichment
function, uses HaemAtlas list included with Cellmix) |
Enrichment of WGCNA modules for a
manually curated list of cell-type specific genes, allowing interpretation of cellular context of the module |
14,
19, Supp.
Ref 30 |
23 | wgcna | cellEx | Hypergeometric test | BH | <0.05 | WGCNA (UserListEnrichment
function, uses Abbas list included with Cellmix |
Enrichment of WGCNA modules for a
manually curated list of cell-type specific genes, allowing interpretation of cellular context of the module |
14,
19, Supp.
Ref 11 |
24 | wgcna | cellprop | Pearson correlation | BH | <0.05 *** | WGCNA (WGCNA::cor function) | Correlation of WGCNA module
eigengenes with estimated cell-type proportions (as output using functions in the CellMix package) |
14 |
25 | CELL | cellEx | mapping | N/A | N/A | Mapping of consensus cell type
names to cell type names used in the BloodLists dataset included with WGCNA) |
Convenience mapping that facilitates
searching for cell types in the ANIMA database regardless of the name of the cell type in the original gene marker list |
14 |
26 | CELL | cellEx | mapping | N/A | N/A | Mapping of consensus cell type
names to cell type names used in the HaemAtlas dataset included with CellMix) |
Convenience mapping that facilitates
searching for cell types in the ANIMA database regardless of the name of the cell type in the original gene marker list |
19, Supp.
Ref 30 |
27 | CELL | cellEx | mapping | N/A | N/A | Mapping of consensus cell type
names to cell type names used in the Abbas dataset included with CellMix) |
Convenience mapping that facilitates
searching for cell types in the ANIMA database regardless of the name of the cell type in the original gene marker list |
19, Supp.
Ref 11 |
28 | CELL | cellprop | mapping | N/A | N/A | Mapping of consensus cell type
names to cell type names used in the Abbas dataset included with CellMix) |
Convenience mapping that facilitates
searching for cell types in the ANIMA database regardless of the name of the cell type in the original gene marker list |
19, Supp.
Ref 11 |
29 | wgcna | pheno | Pearson correlation | BH | <0.05 *** | WGCNA (WGCNA::cor function) | Correlation of WGCNA module
eigengenes with a numeric matrix of clinical and other phenotype data |
14 |
*a minimum of three and a maximum of ten results were included in the ANIMA database
** a minimum of four and a maximum of ten results were included in the ANIMA database
*** results were reviewed algorithmically; in the case of no hits for enrichment at the specified cutoff, less stringent criteria were applied (See Supplementary information). In all cases, the corrected P-value was included as a property of the relationship, allowing filtering of the database entries.
X the PROBE-PROBE network is a monopartite correlation network and not a bipartite graph
Abbreviations: BH, Benjamini-Hochberg multiple testing correction