Contents
library(SummarizedExperiment)
library(dplyr)
library(GenomicSignatures)
library(circlize)
library(grid)
library(tm)
library(SnowballC)
library(wordcloud)
library(RColorBrewer)
library(magrittr)
Load PCAmodels
model_dir = "~/data2/PCAGenomicSignatures_Library"
c2cp = readRDS(file.path(model_dir, "canonicalPathways/recount_canonicalPathways_PCAmodel.rds"))
c6 = readRDS(file.path(model_dir, "oncogenicSignatures/recount_oncogenicSignatures_PCAmodel.rds"))
c7 = readRDS(file.path(model_dir, "immunologicalSignatures/recount_immunologicalSignatures_PCAmodel.rds"))
# load validation datasets
load("~/data2/GenomicSuperSignature/data/TCGA_validationDatasets.rda")
dataset = TCGA_validationDatasets
# color setup
colfunc = colorRampPalette(c("white", "red"))
n = 20
col = colfunc(n)[c(1,2,n)]
Canonical Pathways (c2cp)
Validation
MeSH
drawWordcloud(c2cp, ind = 49, rm.noise = 4, weighted = TRUE)
drawWordcloud(c2cp, ind = 109, rm.noise = 4, weighted = TRUE)
GSEA
subsetPathways(c2cp, c(49, 109))
## DataFrame with 20 rows and 2 columns
## PCcluster49
## <factor>
## Up_1 BIOCARTA_TRKA_PATHWAY
## Up_2 REACTOME_SCAVENGING_OF_HEME_FROM_PLASMA
## Up_3 BIOCARTA_GATA3_PATHWAY
## Up_4 BIOCARTA_MTOR_PATHWAY
## Up_5 BIOCARTA_PYK2_PATHWAY
## ... ...
## Down_6 REACTOME_ORGANELLE_BIOGENESIS_AND_MAINTENANCE
## Down_7 REACTOME_SIGNALING_BY_WNT
## Down_8 REACTOME_MITOTIC_PROMETAPHASE
## Down_9 REACTOME_THE_CITRIC_ACID_TCA_CYCLE_AND_RESPIRATORY_ELECTRON_TRANSPORT
## Down_10 KEGG_WNT_SIGNALING_PATHWAY
## PCcluster109
## <factor>
## Up_1 KEGG_RIBOSOME
## Up_2 REACTOME_NONSENSE_MEDIATED_DECAY_NMD_INDEPENDENT_OF_THE_EXON_JUNCTION_COMPLEX_EJC
## Up_3 REACTOME_SELENOAMINO_ACID_METABOLISM
## Up_4 REACTOME_NONSENSE_MEDIATED_DECAY_NMD
## Up_5 REACTOME_SRP_DEPENDENT_COTRANSLATIONAL_PROTEIN_TARGETING_TO_MEMBRANE
## ... ...
## Down_6 REACTOME_CILIUM_ASSEMBLY
## Down_7 REACTOME_C_TYPE_LECTIN_RECEPTORS_CLRS
## Down_8 REACTOME_MITOTIC_PROMETAPHASE
## Down_9 REACTOME_ESR_MEDIATED_SIGNALING
## Down_10 REACTOME_SIGNALING_BY_HEDGEHOG
Smallest/Largest PCclusters
x = c2cp
indMin = which.min(metadata(x)$size)
indMax = which.max(metadata(x)$size)
metadata(x)$size[indMin]
## [1] 2
metadata(x)$size[indMax]
## [1] 127
drawWordcloud(x, indMin)
drawWordcloud(x, indMax)
Oncogenic Signatures (c6)
Validation
MeSH
drawWordcloud(c6, ind = 68, rm.noise = 4, weighted = TRUE)
drawWordcloud(c6, ind = 136, rm.noise = 4, weighted = TRUE)
GSEA
subsetPathways(c6, c(68, 136))
## DataFrame with 20 rows and 2 columns
## PCcluster68 PCcluster136
## <factor> <factor>
## Up_1 BCAT_BILD_ET_AL_DN PGF_UP.V1_UP
## Up_2 TBK1.DF_DN ERBB2_UP.V1_DN
## Up_3 EIF4E_DN TBK1.DF_DN
## Up_4 PGF_UP.V1_UP VEGF_A_UP.V1_DN
## Up_5 ERBB2_UP.V1_DN RB_P130_DN.V1_DN
## ... ... ...
## Down_6 E2F1_UP.V1_UP ESC_J1_UP_EARLY.V1_DN
## Down_7 STK33_UP MEK_UP.V1_DN
## Down_8 TBK1.DF_UP PRC2_EED_UP.V1_DN
## Down_9 EGFR_UP.V1_DN HOXA9_DN.V1_UP
## Down_10 MEK_UP.V1_DN GCNP_SHH_UP_LATE.V1_UP
Smallest/Largest PCclusters
x = c6
indMin = which.min(metadata(x)$size)
indMax = which.max(metadata(x)$size)
metadata(x)$size[indMin]
## [1] 3
metadata(x)$size[indMax]
## [1] 119
drawWordcloud(x, indMin)
drawWordcloud(x, indMax)
Immunological Signatures (c7)
Validation
MeSH
drawWordcloud(c7, ind = 75, rm.noise = 4, weighted = TRUE)
drawWordcloud(c7, ind = 76, rm.noise = 4, weighted = TRUE)
GESA
subsetPathways(c7, c(75, 76))
## DataFrame with 20 rows and 2 columns
## PCcluster75
## <factor>
## Up_1 GSE2405_0H_VS_24H_A_PHAGOCYTOPHILUM_STIM_NEUTROPHIL_UP
## Up_2 GSE27241_WT_VS_RORGT_KO_TH17_POLARIZED_CD4_TCELL_TREATED_WITH_DIGOXIN_UP
## Up_3 GSE2405_0H_VS_9H_A_PHAGOCYTOPHILUM_STIM_NEUTROPHIL_DN
## Up_4 GSE26030_TH1_VS_TH17_RESTIMULATED_DAY5_POST_POLARIZATION_UP
## Up_5 GSE14000_TRANSLATED_RNA_VS_MRNA_DC_DN
## ... ...
## Down_6 GSE13484_12H_VS_3H_YF17D_VACCINE_STIM_PBMC_UP
## Down_7 GSE17721_PAM3CSK4_VS_CPG_24H_BMDC_DN
## Down_8 GSE45365_WT_VS_IFNAR_KO_CD11B_DC_UP
## Down_9 GSE19888_ADENOSINE_A3R_INH_PRETREAT_AND_ACT_BY_A3R_VS_TCELL_MEMBRANES_ACT_MAST_CELL_UP
## Down_10 GSE17721_12H_VS_24H_GARDIQUIMOD_BMDC_UP
## PCcluster76
## <factor>
## Up_1 GSE9006_TYPE_1_VS_TYPE_2_DIABETES_PBMC_AT_DX_UP
## Up_2 GSE7596_AKT_TRANSD_VS_CTRL_CD4_TCONV_WITH_TGFB_UP
## Up_3 GSE5099_UNSTIM_VS_MCSF_TREATED_MONOCYTE_DAY7_UP
## Up_4 GSE6269_HEALTHY_VS_STAPH_AUREUS_INF_PBMC_UP
## Up_5 GSE37416_CTRL_VS_6H_F_TULARENSIS_LVS_NEUTROPHIL_DN
## ... ...
## Down_6 GSE27786_CD8_TCELL_VS_NKTCELL_DN
## Down_7 GSE3982_MAST_CELL_VS_NEUTROPHIL_UP
## Down_8 GSE13484_12H_VS_3H_YF17D_VACCINE_STIM_PBMC_DN
## Down_9 GSE2770_IL4_ACT_VS_ACT_CD4_TCELL_2H_DN
## Down_10 GSE411_UNSTIM_VS_100MIN_IL6_STIM_SOCS3_KO_MACROPHAGE_DN
Smallest/Largest PCclusters
x = c7
indMin = which.min(metadata(x)$size)
indMax = which.max(metadata(x)$size)
metadata(x)$size[indMin]
## [1] 3
metadata(x)$size[indMax]
## [1] 229
drawWordcloud(x, indMin)
drawWordcloud(x, indMax)
