leading edge analysis

leading edge and core enrichment

Leading edge analysis reports Tags to indicate the percentage of genes contributing to the enrichment score, List to indicate where in the list the enrichment score is attained and Signal for enrichment signal strength.

It would also be very interesting to get the core enriched genes that contribute to the enrichment.

Now DOSE, clusterProfiler and ReactomePA all support leading edge analysis and report core enriched genes.

library(DOSE)
data(geneList)
x <- gseDO(geneList)

## [1] "preparing geneSet collections..."
## [1] "GSEA analysis..."
## [1] "leading edge analysis..."
## [1] "done..."

head(x)

##                        ID                             Description setSize
## DOID:1492       DOID:1492                  eye and adnexa disease     459
## DOID:5614       DOID:5614                             eye disease     450
## DOID:0060037 DOID:0060037 developmental disorder of mental health     315
## DOID:374         DOID:374                       nutrition disease     313
## DOID:5679       DOID:5679                         retinal disease     299
## DOID:654         DOID:654                           overnutrition     298
##              enrichmentScore       NES      pvalue   p.adjust    qvalues
## DOID:1492         -0.3105160 -1.379586 0.001231527 0.03329506 0.02257366
## DOID:5614         -0.3125247 -1.386602 0.001243781 0.03329506 0.02257366
## DOID:0060037      -0.3407349 -1.473028 0.001307190 0.03329506 0.02257366
## DOID:374          -0.3421127 -1.477222 0.001310616 0.03329506 0.02257366
## DOID:5679         -0.3676313 -1.582707 0.001333333 0.03329506 0.02257366
## DOID:654          -0.3490198 -1.500301 0.001340483 0.03329506 0.02257366
##              rank                   leading_edge
## DOID:1492    1793 tags=22%, list=14%, signal=19%
## DOID:5614    1768 tags=22%, list=14%, signal=19%
## DOID:0060037 2313 tags=24%, list=19%, signal=20%
## DOID:374     1464 tags=22%, list=12%, signal=20%
## DOID:5679    1768 tags=24%, list=14%, signal=21%
## DOID:654     1464 tags=21%, list=12%, signal=19%
##                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         core_enrichment
## DOID:1492    3371/3082/5914/2878/4153/3791/23247/1543/80184/6750/1958/2098/7450/596/9187/2034/482/948/1490/1280/3931/5737/4314/4881/2261/3426/187/629/6403/7042/6785/7507/2934/5176/4060/1277/7078/5950/2057/727/10516/4311/2247/1295/358/10203/2192/582/10218/57125/3485/585/1675/6310/2202/4313/2944/4254/3075/1501/2099/3480/4653/1195/6387/3305/1471/857/4016/1909/4053/6678/1296/7033/4915/55812/1191/5654/10631/2152/2697/7043/2952/6935/2200/3572/7177/7031/3479/2006/10451/9370/771/3117/125/652/4693/5346/1524
## DOID:5614              3082/5914/2878/4153/3791/23247/1543/80184/6750/1958/2098/7450/596/9187/2034/482/948/1490/1280/3931/5737/4314/4881/2261/3426/187/629/6403/7042/6785/7507/2934/5176/4060/1277/7078/5950/2057/727/10516/4311/2247/1295/358/10203/2192/582/10218/57125/3485/585/1675/6310/2202/4313/2944/4254/3075/1501/2099/3480/4653/6387/3305/1471/857/4016/1909/4053/6678/1296/7033/4915/55812/1191/5654/10631/2152/2697/7043/2952/6935/2200/3572/7177/7031/3479/2006/10451/9370/771/3117/125/652/4693/5346/1524
## DOID:0060037                                                                                                         154/1760/9732/7337/5175/6532/4763/54806/9759/6326/1499/7157/221037/627/2908/3399/2571/3082/23503/3791/51265/27347/55650/596/3067/51185/7552/22829/23426/324/5021/4885/7248/8910/8604/3397/4208/3400/26470/553/3953/6812/64221/80208/5172/9037/3952/477/93664/3625/2944/6925/6594/7102/3908/2550/4915/4922/26960/1746/2697/6863/3913/2891/367/4128/150/7166/6505/5348/18/4129/9370/57502/4137/79083
## DOID:374                                                                                                                                                                    2169/1490/7840/4887/4314/595/4018/6403/590/3087/866/66036/5919/5176/3953/164656/5950/2638/2166/5243/5468/5108/10560/4023/3485/7350/3952/1149/585/1513/3489/79068/4671/477/4313/3625/9369/6720/7494/2099/3480/3991/23446/6678/4915/5167/8228/165/2152/185/367/4982/3667/4128/9607/3572/150/563/1489/3479/9370/9122/5105/2167/5346/79689/5241
## DOID:5679                                                                                                                                          2878/3791/23247/80184/6750/7450/596/9187/2034/482/948/1490/1280/5737/4314/4881/3426/187/629/6403/6785/2934/5176/7078/5950/727/10516/4311/2247/1295/358/10203/582/10218/57125/585/1675/6310/2202/4313/2944/4254/3075/2099/3480/4653/6387/1471/857/4016/1909/4053/6678/1296/4915/55812/1191/5654/10631/2697/2952/6935/2200/3479/2006/10451/9370/771/652/4693/5346/1524
## DOID:654                                                                                                                                                                                            2169/1490/7840/4887/4314/595/4018/6403/590/3087/866/66036/5919/5176/3953/5950/2638/2166/5243/5468/4023/3485/7350/3952/1149/585/1513/3489/79068/4671/477/4313/3625/9369/6720/7494/2099/3480/3991/6678/4915/5167/8228/165/2152/185/367/4982/3667/4128/9607/3572/150/563/1489/3479/9370/9122/5105/2167/5346/79689/5241

cnetplot for core enriched genes

With core enriched genes reported, cnetplot function for GSEA result is also implemented.

setReadable

To make the plot more readable, we can use setReadable to convert Entrez ID to Gene Symbol.

y <- setReadable(x, 'org.Hs.eg.db')

head(y, 2)

##                        ID                             Description setSize
## DOID:1492       DOID:1492                  eye and adnexa disease     459
## DOID:5614       DOID:5614                             eye disease     450
##              enrichmentScore       NES      pvalue   p.adjust    qvalues
## DOID:1492         -0.3105160 -1.379586 0.001231527 0.03329506 0.02257366
## DOID:5614         -0.3125247 -1.386602 0.001243781 0.03329506 0.02257366
##              rank                   leading_edge
## DOID:1492    1793 tags=22%, list=14%, signal=19%
## DOID:5614    1768 tags=22%, list=14%, signal=19%
##                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       core_enrichment
## DOID:1492    TNC/HGF/RARA/GPX3/MBL2/KDR/KIAA0556/CYP1A1/CEP290/SST/EGR1/ESD/VWF/BCL2/SLC24A1/EPAS1/ATP1B2/CD36/CTGF/COL2A1/LCAT/PTGFR/MMP3/NPR1/FGFR3/CFI/APLNR/CFB/SELP/TGFB2/ELOVL4/XPA/GSN/SERPINF1/LUM/COL1A1/TIMP3/RBP4/EPOR/C5/FBLN5/MME/FGF2/COL8A1/AQP1/CALCRL/FBLN1/BBS1/ANGPTL7/PLXDC1/IGFBP2/BBS4/CFD/ATXN1/EFEMP1/MMP2/GSTM1/KITLG/CFH/CTNND2/ESR1/IGF1R/MYOC/CLK1/CXCL12/HSPA1L/CST3/CAV1/LOXL1/EDNRA/LTBP2/SPARC/COL8A2/TFF3/NTRK2/SPATA7/CLU/HTRA1/POSTN/F3/GJA1/TGFB3/GSTT1/ZEB1/FBN1/IL6ST/TPSAB1/TFF1/IGF1/ELN/VAV3/ADIPOQ/CA12/HLA-DQA1/ADH1B/BMP4/NDP/PLIN1/CX3CR1
## DOID:5614             HGF/RARA/GPX3/MBL2/KDR/KIAA0556/CYP1A1/CEP290/SST/EGR1/ESD/VWF/BCL2/SLC24A1/EPAS1/ATP1B2/CD36/CTGF/COL2A1/LCAT/PTGFR/MMP3/NPR1/FGFR3/CFI/APLNR/CFB/SELP/TGFB2/ELOVL4/XPA/GSN/SERPINF1/LUM/COL1A1/TIMP3/RBP4/EPOR/C5/FBLN5/MME/FGF2/COL8A1/AQP1/CALCRL/FBLN1/BBS1/ANGPTL7/PLXDC1/IGFBP2/BBS4/CFD/ATXN1/EFEMP1/MMP2/GSTM1/KITLG/CFH/CTNND2/ESR1/IGF1R/MYOC/CXCL12/HSPA1L/CST3/CAV1/LOXL1/EDNRA/LTBP2/SPARC/COL8A2/TFF3/NTRK2/SPATA7/CLU/HTRA1/POSTN/F3/GJA1/TGFB3/GSTT1/ZEB1/FBN1/IL6ST/TPSAB1/TFF1/IGF1/ELN/VAV3/ADIPOQ/CA12/HLA-DQA1/ADH1B/BMP4/NDP/PLIN1/CX3CR1

cnetplot

cnetplot(y, foldChange=geneList)

In the figure, we can find that most significant associated diseases are down-regulated.

We can of course visualize significant up-regulated disease pathways.

up <- y$Description[order(y$NES, decreasing=TRUE)][1:6]
up

## [1] "tuberculosis"                                  
## [2] "respiratory syncytial virus infectious disease"
## [3] "laryngeal carcinoma"                           
## [4] "Kawasaki disease"                              
## [5] "breast ductal carcinoma"                       
## [6] "pulmonary sarcoidosis"

cnetplot(y, foldChange=geneList, showCategory=up)

gseaplot

gseaplot now supports adding plot title.

gseaplot(y, y$ID[1], title=y$Description[1])

Citation

G Yu, LG Wang, GR Yan, QY He. DOSE: an R/Bioconductor package for Disease Ontology Semantic and Enrichment analysis. Bioinformatics 2015, 31(4):608-609.

comments powered by Disqus