数据分析：基于DESeq2的转录组功能富集分析

介绍

DESeq2常用于识别差异基因，它主要使用了标准化因子标准化数据，再根据广义线性模型判别组间差异（组间残差是否显著判断）。在获取差异基因结果后，我们可以进行下一步的富集分析，常用方法有基于在线网站DAVID以及脚本处理的两类，本文介绍基于fgsea的方法计算富集分析得分。

DESeq2差异分析

了解DESeq2如何标准化数据和识别差异基因。下面给出简要代码

library(DESeq2)
library(airway)
data("airway")
ddsSE <- DESeqDataSet(airway, design = ~ cell + dex)
ddsSE <- DESeq(ddsSE)
res <- results(ddsSE, tidy = TRUE) %>% na.omit() %>% as_tibble()

head(res)

# A tibble: 6 x 7
  row             baseMean log2FoldChange  lfcSE   stat     pvalue      padj
  <chr>              <dbl>          <dbl>  <dbl>  <dbl>      <dbl>     <dbl>
1 ENSG00000000003    709.          0.381  0.101   3.79  0.000152   0.00128  
2 ENSG00000000419    520.         -0.207  0.112  -1.84  0.0653     0.197    
3 ENSG00000000457    237.         -0.0379 0.143  -0.264 0.792      0.911    
4 ENSG00000000460     57.9         0.0882 0.287   0.307 0.759      0.895    
5 ENSG00000000971   5817.         -0.426  0.0883 -4.83  0.00000138 0.0000182
6 ENSG00000001036   1282.          0.241  0.0887  2.72  0.00658    0.0328

转换geneID

我们使用的MSigDB数据库的pathway 基因ID只有entrez和HGNC symbol两类，如果是ensemble id，需要转换。

library(org.Hs.eg.db)
library(tidyverse)
ens2symbol <- AnnotationDbi::select(org.Hs.eg.db,
                                    key=res$row, 
                                    columns="SYMBOL",
                                    keytype="ENSEMBL")
ens2symbol <- as_tibble(ens2symbol)
head(ens2symbol)

# A tibble: 6 x 2
  ENSEMBL         SYMBOL  
  <chr>           <chr>   
1 ENSG00000000003 TSPAN6  
2 ENSG00000000419 DPM1    
3 ENSG00000000457 SCYL3   
4 ENSG00000000460 C1orf112
5 ENSG00000000971 CFH     
6 ENSG00000001036 FUCA2

合并数据；过滤NA值；去重；重复基因求stat（stat数据作为排序指标用于后续富集分析）

res2 <- inner_join(res, ens2symbol, by=c("row"="ENSEMBL")) %>% 
  dplyr::select(SYMBOL, stat) %>% 
  na.omit() %>% 
  distinct() %>% 
  group_by(SYMBOL) %>% 
  summarize(stat=mean(stat))
head(res2 )

# A tibble: 6 x 2
  SYMBOL       stat
  <chr>       <dbl>
1 A1BG      0.680  
2 A1BG-AS1 -1.79   
3 A2M      -1.26   
4 A2M-AS1   0.875  
5 A4GALT   -4.14   
6 A4GNT     0.00777

构建fgsea输入数据

基因排序值转换

library(fgsea)

ranks <- deframe(res2)
head(ranks, 20)

        A1BG     A1BG-AS1          A2M      A2M-AS1       A4GALT        A4GNT         AAAS         AACS 
 0.679946437 -1.793291412 -1.259539478  0.875346116 -4.144839902  0.007772497  0.163986128  1.416071728 
     AADACL4        AADAT        AAGAB         AAK1        AAMDC         AAMP         AAR2        AARS1 
-1.876311694  3.079128034  1.554279946  1.141522348 -2.147527241 -3.170612332 -2.364380163  4.495474603 
       AARS2       AARSD1        AASDH     AASDHPPT 
 5.057470292  0.654208006  0.665531695 -0.353496148

pathways的基因集合，上MSigDB下载基因集。演示使用KEGG基因集

pathways.hallmark <- gmtPathways("../../Result/GeneID/msigdb.v7.1.symbols_KEGG.gmt")
pathways.hallmark %>% 
  head() %>% 
  lapply(head)

$KEGG_GLYCOLYSIS_GLUCONEOGENESIS
[1] "ACSS2" "GCK"   "PGK2"  "PGK1"  "PDHB"  "PDHA1"

$KEGG_CITRATE_CYCLE_TCA_CYCLE
[1] "IDH3B" "DLST"  "PCK2"  "CS"    "PDHB"  "PCK1" 

$KEGG_PENTOSE_PHOSPHATE_PATHWAY
[1] "RPE"   "RPIA"  "PGM2"  "PGLS"  "PRPS2" "FBP2" 

$KEGG_PENTOSE_AND_GLUCURONATE_INTERCONVERSIONS
[1] "UGT1A10" "UGT1A8"  "RPE"     "UGT1A7"  "UGT1A6"  "UGT2B28"

$KEGG_FRUCTOSE_AND_MANNOSE_METABOLISM
[1] "MPI"  "PMM2" "PMM1" "FBP2" "PFKM" "GMDS"

$KEGG_GALACTOSE_METABOLISM
[1] "GCK"     "GALK1"   "GLB1"    "GALE"    "B4GALT1" "PGM2"

运行

fgseaRes <- fgsea(pathways=pathways.hallmark, stats=ranks, nperm=1000)
head(fgseaRes[order(pval), ])

从查看KEGG_REGULATION_OF_ACTIN_CYTOSKELETON富集分数分布

plotEnrichment(pathways.hallmark[["KEGG_REGULATION_OF_ACTIN_CYTOSKELETON"]],
               ranks) + labs(title="KEGG_REGULATION_OF_ACTIN_CYTOSKELETON")

查看上下调通路结果

topPathwaysUp <- fgseaRes[ES > 0][head(order(pval), n=10), pathway]
topPathwaysDown <- fgseaRes[ES < 0][head(order(pval), n=10), pathway]
topPathways <- c(topPathwaysUp, rev(topPathwaysDown))
plotGseaTable(pathways.hallmark[topPathways], ranks, fgseaRes, 
              gseaParam=0.5)

其他展示方式

fgseaResTidy <- fgseaRes %>%
  as_tibble() %>%
  arrange(desc(NES))

# Show in a nice table:
fgseaResTidy %>% 
  dplyr::select(-leadingEdge, -ES, -nMoreExtreme) %>% 
  arrange(padj) %>% 
  DT::datatable()

ggplot(fgseaResTidy, aes(reorder(pathway, NES), NES)) +
  geom_col(aes(fill = padj<0.0001)) +
  coord_flip() +
  labs(x="Pathway", y="Normalized Enrichment Score",
       title="Hallmark pathways NES from GSEA") + 
  theme_minimal()

查看通路的基因

res_temp <- inner_join(res, ens2symbol, by=c("row"="ENSEMBL"))
pathways.hallmark %>% 
  enframe("pathway", "SYMBOL") %>% 
  unnest(cols = c(SYMBOL)) %>% 
  inner_join(res_temp , by="SYMBOL") %>%
  head()

# A tibble: 6 x 9
  pathway                         SYMBOL row             baseMean log2FoldChange lfcSE   stat pvalue   padj
  <chr>                           <chr>  <chr>              <dbl>          <dbl> <dbl>  <dbl>  <dbl>  <dbl>
1 KEGG_GLYCOLYSIS_GLUCONEOGENESIS ACSS2  ENSG00000131069    669.         -0.269  0.114 -2.35  0.0188 0.0756
2 KEGG_GLYCOLYSIS_GLUCONEOGENESIS GCK    ENSG00000106633     28.8         0.305  0.374  0.815 0.415  0.662 
3 KEGG_GLYCOLYSIS_GLUCONEOGENESIS PGK1   ENSG00000102144   7879.         -0.300  0.353 -0.850 0.395  0.642 
4 KEGG_GLYCOLYSIS_GLUCONEOGENESIS PDHB   ENSG00000168291    648.         -0.257  0.102 -2.52  0.0117 0.0521
5 KEGG_GLYCOLYSIS_GLUCONEOGENESIS PDHA1  ENSG00000131828    651.         -0.0744 0.104 -0.715 0.475  0.710 
6 KEGG_GLYCOLYSIS_GLUCONEOGENESIS PGM2   ENSG00000169299    302.         -0.315  0.136 -2.33  0.0201 0.0797

其他用法

miR targets

fgsea(pathways=gmtPathways("msigdb/c3.mir.v6.2.symbols.gmt"), ranks, nperm=1000) %>% 
  as_tibble() %>% 
  arrange(padj)

GO annotations

fgsea(pathways=gmtPathways("msigdb/c5.all.v6.2.symbols.gmt"), ranks, nperm=1000) %>% 
  as_tibble() %>% 
  arrange(padj)

非人物种

library(biomaRt)
mart <- useDataset("mmusculus_gene_ensembl", mart=useMart("ensembl"))
bm <- getBM(attributes=c("ensembl_gene_id", "hsapiens_homolog_associated_gene_name"), mart=mart) %>%
  distinct() %>%
  as_tibble() %>%
  na_if("") %>% 
  na.omit()
bm