Error with merge_result with dataframes using clusterprofiler

Question

[Vasu]

I'm using clusterprofiler for pathway analysis for two datasets and want to plot the results both in same plot. I have seen an example to merge the results here from GuangchuangYu github

But here it is merging of enrichmentesults. In my case I have results in dataframes. Below I gave a n example of two dataframes A4 and B4

head(A4)

                             ID                   Description GeneRatio  BgRatio
1         HALLMARK_ADIPOGENESIS         HALLMARK_ADIPOGENESIS     10/60 200/4386
2  HALLMARK_ALLOGRAFT_REJECTION  HALLMARK_ALLOGRAFT_REJECTION     33/55 200/4386
3         HALLMARK_ANGIOGENESIS         HALLMARK_ANGIOGENESIS     9/105  36/4386
4      HALLMARK_APICAL_JUNCTION      HALLMARK_APICAL_JUNCTION       3/9 200/4386
5            HALLMARK_APOPTOSIS            HALLMARK_APOPTOSIS      7/54 161/4386
6 HALLMARK_BILE_ACID_METABOLISM HALLMARK_BILE_ACID_METABOLISM      3/10 112/4386
        pvalue     p.adjust       qvalue
1 3.172592e-04 4.230123e-03 3.784847e-03
2 2.161865e-31 3.675170e-30 2.275647e-30
3 1.005961e-07 1.911325e-06 1.694250e-06
4 6.398199e-03 2.559279e-02 2.020484e-02
5 3.261885e-03 3.914262e-02 3.204659e-02
6 1.706706e-03 2.218718e-02 1.616879e-02
                                                                                                                                                                                    geneID
1                                                                                                                            ALDOA/ANGPTL4/CMPK1/CYP4B1/ETFB/HSPB8/PPM1B/RIOK3/SLC19A1/UBC
2 CCL19/CCR2/CD2/CD247/CD28/CD3D/CD3E/CD3G/CD40/CD40LG/CD8A/CD96/CRTAM/CXCL13/CXCL9/CXCR3/FASLG/GZMA/GZMB/IL12RB1/IL16/IL18RAP/IL2RB/IL2RG/ITGAL/ITK/LCK/PRF1/PRKCB/SIT1/STAT4/TRAT1/ZAP70
3                                                                                                                                    CCND2/COL3A1/COL5A2/FSTL1/LUM/NRP1/POSTN/SLCO2A1/VCAN
4                                                                                                                                                                        DHX16/MAPK13/PBX2
5                                                                                                                                           CAV1/CDKN1A/FEZ1/GADD45A/IGFBP6/PLAT/TNFRSF12A
6                                                                                                                                                                     PEX6/SLC29A1/SLC35B2
  Count   
1    10        
2    33             
3     9        
4     3 
5     7            
6     3

head(B4)

                            ID                  Description GeneRatio  BgRatio
1        HALLMARK_ADIPOGENESIS        HALLMARK_ADIPOGENESIS      3/10 200/4386
2 HALLMARK_ALLOGRAFT_REJECTION HALLMARK_ALLOGRAFT_REJECTION     16/54 200/4386
3        HALLMARK_ANGIOGENESIS        HALLMARK_ANGIOGENESIS      4/99  36/4386
4     HALLMARK_APICAL_JUNCTION     HALLMARK_APICAL_JUNCTION       2/5 200/4386
5         HALLMARK_COAGULATION         HALLMARK_COAGULATION      8/70 138/4386
6          HALLMARK_COMPLEMENT          HALLMARK_COMPLEMENT     13/54 200/4386
        pvalue     p.adjust       qvalue
1 8.836224e-03 4.418112e-02 3.720515e-02
2 8.785433e-10 1.186033e-08 8.323041e-09
3 8.268709e-03 6.118844e-02 4.874186e-02
4 1.888702e-02 5.666106e-02 3.976215e-02
5 1.425831e-03 8.554987e-03 6.003500e-03
6 5.303804e-07 3.580067e-06 2.512328e-06
                                                                                    geneID
1                                                                        CD151/PFKL/TALDO1
2 CD1D/CD28/CD7/CD80/CD86/CSF1/CTSS/CXCL9/FCGR2B/GPR65/HLA-DMA/IGSF6/IL2RA/ITGB2/LCK/NLRP3
3                                                                     FSTL1/MSX1/NRP1/VCAN
4                                                                               CRAT/GNAI2
5                                                    ACOX2/C1R/C1S/C3/CTSB/F10/F3/SERPINA1
6        APOBEC3G/CASP1/CASP10/CTSS/DOCK10/GP1BA/IRF1/ITGAM/LCK/OLR1/PIK3CG/PIK3R5/TNFAIP3
  Count   
1     3        
2    16             
3     4        
4     2 
5     8             
6    13

I tried using the merge_result function but it gave me an error.

merge_result(list(Dataset1=A4, Dataset2=B4)) %>%
    plot(., showCategory=10)

Error in as.double(y) : 
  cannot coerce type 'S4' to vector of type 'double'

Is there a way to convert dataframe into enrichmentresult for plotting?

[Santosh Anand]

How did you obtain A4 & B4? What does class(A4) and class(B4) return?

[Vasu]

After using enrichr function from clusterprofiler I saved the result in a dataframe.

class(A4) and class(B4) is "data.frame".

But I see that merge_result function from clusterprofiler uses only enrichmentresult for plotting.

[Santosh Anand]

So you got the answer - save them as original enrichmentresult object in a variable without coercing them as data.frame

[Vasu]

Coercing a dataframe results again into enrichmentresult object is not possible I guess.

But Anyways from GuangchuangYu github I see an example showing Upregulated and downregulated dotplots side by side. So, I followed that.

First I created a new column Dataset in both dataframes A4 and B4.

A4$Dataset <- "Dataset1"
B4$Dataset <- "Dataset2"

I merged both the dataframes.

D <- merge(A4,B4, all=TRUE)

With the code mentioned in above link I used ggplot.

p <- ggplot(D, aes(x = Count, y = fct_reorder(ID, Count))) + 
  geom_point(aes(size = Count, color = p.adjust)) +
  theme_bw(base_size = 14) +
  scale_colour_gradient(limits=c(0, 0.10), low="red") +
  ylab(NULL) +
  ggtitle("Hallmark Genesets")

p + facet_grid(.~Dataset)

I got the plot like below.