Hello friends,

I'm trying to use Pathview for pathway analysis of my differentially expressed genes. I used the below command

pathview(gene.data=foldchanges, pathway.id=keggresids, species="hsa", limit = c(-10,10))

before this I run gage function to get up-regulated pathways from my all differentially expressed genes. Then, I extracted kegg ids and gave as input to pathview function.

keggres = gage(foldchanges, gsets=kegg.sets.hs, same.dir=TRUE)
keggrespathways <- rownames(keggres$greater)
keggresids = substr(keggrespathways, start=1, stop=8)

Finally, I got KEGG pathway plots that have both red and green colors. My question is : 1) If it is up-regulated pathway, why do I get both red and green colors. I know that red color indicates positive fold value, so I thought I should get mostly red colored genes for up-regulated pathway.

2) Also, in this keggres$greater, greater than what? I do not clearly understand the meaning of up-regulated pathway. Tutorial says that gage performs some pair-wise comparisons for all proper gene sets, and I do not get the meaning.

Here is my R session info :

> sessionInfo()
R version 3.6.2 (2019-12-12)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows >= 8 x64 (build 9200)

Matrix products: default

locale:
[1] LC_COLLATE=Korean_Korea.949  LC_CTYPE=Korean_Korea.949   
[3] LC_MONETARY=Korean_Korea.949 LC_NUMERIC=C                
[5] LC_TIME=Korean_Korea.949    

attached base packages:
[1] parallel  stats4    stats     graphics  grDevices utils     datasets 
[8] methods   base     

other attached packages:
 [1] gageData_2.24.0             gage_2.36.0                
 [3] pathview_1.26.0             org.Hs.eg.db_3.10.0        
 [5] AnnotationDbi_1.48.0        DESeq2_1.26.0              
 [7] SummarizedExperiment_1.16.1 DelayedArray_0.12.2        
 [9] BiocParallel_1.20.1         matrixStats_0.55.0         
[11] Biobase_2.46.0              GenomicRanges_1.38.0       
[13] GenomeInfoDb_1.22.0         IRanges_2.20.2             
[15] S4Vectors_0.24.3            BiocGenerics_0.32.0        

loaded via a namespace (and not attached):
 [1] httr_1.4.1             bit64_0.9-7            splines_3.6.2         
 [4] Formula_1.2-3          assertthat_0.2.1       latticeExtra_0.6-29   
 [7] blob_1.2.1             GenomeInfoDbData_1.2.2 yaml_2.2.0            
[10] pillar_1.4.3           RSQLite_2.2.0          backports_1.1.5       
[13] lattice_0.20-38        glue_1.3.2             digest_0.6.25         
[16] RColorBrewer_1.1-2     XVector_0.26.0         checkmate_2.0.0       
[19] colorspace_1.4-1       htmltools_0.4.0        Matrix_1.2-18         
[22] XML_3.99-0.3           pkgconfig_2.0.3        genefilter_1.68.0     
[25] zlibbioc_1.32.0        purrr_0.3.3            xtable_1.8-4          
[28] scales_1.1.0           jpeg_0.1-8.1           htmlTable_1.13.3      
[31] tibble_2.1.3           annotate_1.64.0        KEGGREST_1.26.1       
[34] ggplot2_3.3.0          nnet_7.3-12            survival_3.1-8        
[37] magrittr_1.5           crayon_1.3.4           KEGGgraph_1.46.0      
[40] memoise_1.1.0          foreign_0.8-72         graph_1.64.0          
[43] tools_3.6.2            data.table_1.12.8      lifecycle_0.1.0       
[46] stringr_1.4.0          locfit_1.5-9.1         munsell_0.5.0         
[49] cluster_2.1.0          Biostrings_2.54.0      compiler_3.6.2        
[52] rlang_0.4.4            grid_3.6.2             RCurl_1.98-1.1        
[55] rstudioapi_0.11        htmlwidgets_1.5.1      bitops_1.0-6          
[58] base64enc_0.1-3        gtable_0.3.0           DBI_1.1.0             
[61] R6_2.4.1               gridExtra_2.3          knitr_1.28            
[64] dplyr_0.8.4            bit_1.1-15.2           Hmisc_4.3-1           
[67] Rgraphviz_2.30.0       stringi_1.4.6          Rcpp_1.0.4            
[70] geneplotter_1.64.0     vctrs_0.2.3            rpart_4.1-15          
[73] acepack_1.4.1          png_0.1-7              tidyselect_1.0.0      
[76] xfun_0.12

Please kindly let me know some answers. Thank you.

For any gene set test, the goal is to say if the gene set in aggregate is up or down-regulated. This is different from saying that every single gene in the pathway is up or down. Depending on the test (and here I am not talking about gage in particular) you have to have some measure of the aggregate signal from a gene set, and how you interpret that is dependent on what the statistic is.

In general I would caution against using a method blindly, without understanding what is happening. How else can you explain to others what you have done?