Enrichment analysis of tissue-specific genes in MHBs

library(rGREAT)
library("org.Hs.eg.db")
library(pheatmap)
options(warn=-1)

An example using colon MHBs

gr=read.table("colon_MHB.bed")
colnames(gr)=c("chr","start","end")
gr=GRanges(gr)
start(gr) = start(gr) + 1
head(gr)

GRanges object with 6 ranges and 0 metadata columns:
      seqnames        ranges strand
         <Rle>     <IRanges>  <Rle>
  [1]     chr1 540839-540886      *
  [2]     chr1 564471-564534      *
  [3]     chr1 566287-566317      *
  [4]     chr1 567312-567438      *
  [5]     chr1 567681-567786      *
  [6]     chr1 568529-568570      *
  -------
  seqinfo: 24 sequences from an unspecified genome; no seqlengths

# loading tissue-specific genes
TS = readRDS("tissueSpecificGenes.RDS")
ls(TS)

1. 'Adipose Tissue'
2. 'Adrenal Gland'
3. 'Appendix'
4. 'Bone Marrow'
5. 'Breast'
6. 'Cerebral Cortex'
7. 'Cervix, uterine'
8. 'Colon'
9. 'Duodenum'
10. 'Endometrium'
11. 'Esophagus'
12. 'Fallopian Tube'
13. 'Gallbladder'
14. 'Heart Muscle'
15. 'Kidney'
16. 'Liver'
17. 'Lung'
18. 'Lymph Node'
19. 'Ovary'
20. 'Pancreas'
21. 'Placenta'
22. 'Prostate'
23. 'Rectum'
24. 'Salivary Gland'
25. 'Seminal Vesicle'
26. 'Skeletal Muscle'
27. 'Skin'
28. 'Small Intestine'
29. 'Smooth Muscle'
30. 'Spleen'
31. 'Stomach'
32. 'Testis'
33. 'Thyroid Gland'
34. 'Tonsil'
35. 'Urinary Bladder'

# colon-specific genes
example = as.character(TS[[8]][,1])
head(example)

1. 'PIGY'
2. 'NOS2'
3. 'CEACAM7'
4. 'NOX1'
5. 'CLCA1'
6. 'CLCA4'

SR = unique(mapIds(org.Hs.eg.db, example, 'ENTREZID', 'SYMBOL'))
SR = SR[!is.na(SR)]
gs = list(example = SR)

res = great(gr, gs, "hg19", verbose = FALSE)
tb = getEnrichmentTable(res)

'select()' returned 1:1 mapping between keys and columns

* extended_tss is already cached, directly use it.

# enrichment of colon-specific genes in colon MHBs with high significance
tb

A data.frame: 1 × 12

	id	genome_fraction	observed_region_hits	fold_enrichment	p_value	p_adjust	mean_tss_dist	observed_gene_hits	gene_set_size	fold_enrichment_hyper	p_value_hyper	p_adjust_hyper
	<chr>	<dbl>	<int>	<dbl>	<dbl>	<dbl>	<dbl>	<int>	<int>	<dbl>	<dbl>	<dbl>
1	example	0.009976902	208	1.438598	3.721069e-07	3.721069e-07	48792	70	143	1.109497	0.1394002	0.1394002

test=tb[,c('id','p_adjust')]
test$rnkPV=rownames(test)
log=test$p_adjust>0.05
test$rnkPV[log]<-NA
test

A data.frame: 1 × 3

	id	p_adjust	rnkPV
	<chr>	<dbl>	<chr>
1	example	3.721069e-07	1

Show enrichment by ranks of significance

result=read.table("Exp.txt",sep="\t",header = T, row.names = 1)

head(result)

A data.frame: 6 × 17

	Adipose	Adrenal	B.cell	Breast	colon	esophagus	heart	liver	lung	ovary	pancreas	placenta	spleen	stomach	T.cell	thymus	thyroid
	<int>	<int>	<lgl>	<int>	<int>	<int>	<int>	<int>	<int>	<int>	<int>	<int>	<int>	<int>	<lgl>	<int>	<int>
Adipose Tissue	1	NA	NA	5	NA	NA	2	NA	NA	NA	NA	3	NA	NA	NA	NA	NA
Adrenal Gland	NA	1	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	2	NA
Breast	NA	NA	NA	1	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA
Colon	NA	NA	NA	NA	1	NA	NA	NA	NA	NA	2	NA	NA	NA	NA	NA	NA
Esophagus	NA	NA	NA	3	NA	1	NA	NA	NA	NA	NA	2	NA	2	NA	NA	NA
Heart Muscle	NA	NA	NA	NA	NA	NA	1	NA	NA	NA	NA	NA	NA	NA	NA	NA	NA

# mask non-significant ones
result_input=result
result_input[is.na(result_input)] <- 14
head(result_input)

A data.frame: 6 × 17

	Adipose	Adrenal	B.cell	Breast	colon	esophagus	heart	liver	lung	ovary	pancreas	placenta	spleen	stomach	T.cell	thymus	thyroid
	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>
Adipose Tissue	1	14	14	5	14	14	2	14	14	14	14	3	14	14	14	14	14
Adrenal Gland	14	1	14	14	14	14	14	14	14	14	14	14	14	14	14	2	14
Breast	14	14	14	1	14	14	14	14	14	14	14	14	14	14	14	14	14
Colon	14	14	14	14	1	14	14	14	14	14	2	14	14	14	14	14	14
Esophagus	14	14	14	3	14	1	14	14	14	14	14	2	14	2	14	14	14
Heart Muscle	14	14	14	14	14	14	1	14	14	14	14	14	14	14	14	14	14

Heatmap

options(repr.plot.width = 8, repr.plot.height=6, repr.plot.res = 150)
pheatmap(result_input,
    show_rownames=T,main="Tissue-specific gene",
    cluster_cols= F,angle_col=45,fontsize_row=11,cellwidth=21,
    cluster_rows=F,cellheight=18,fontsize_col=12,how_colnames = T,
    legend_breaks = c(1,14),legend_labels = c("1","14"),
    scale = "none",border_color = "grey90", display_numbers = matrix(ifelse(result_input > 1, "", "*"), nrow = nrow(result_input)),
    color = colorRampPalette(colors = c("#F71E35", "#FCD271","#fcebb6","white"))(50))