Diseases related pathways

Question

[mrashad]

I have multiple genes collected from diabetic patients, I need to find out the pathways related to these genes using KEGG, Reactome and Wikipathways. I know how to do that but need to find out if the resulted pathways are related to Diabetes or not. Please help me to do that, is there a tool or R package help me to find out diabetes pathways ?

Answer 1

[rpolicastro]

The Molecular Signatures Database from Broad/UCSD has diabetes specific terms. Just search for 'diabetes' here to see what is included. You'll notice in the results that it also aggregates terms from other databases such as KEGG, Reactome, and Wikipathway as well.

[i.sudbery]

This is probably the best you can do with databases. But nothing is really going to beat either a) Learning enough about the biology of diabetes that you know if the pathways are involved or b) showing your list of pathways to someone who has an indepth knowledge of Diabetes biology.

[mrashad]

What is the difference between KEGG diabetes here: https://www.gsea-msigdb.org/gsea/msigdb/cards/KEGG_TYPE_II_DIABETES_MELLITUS.html and KEGG diabetes here: https://www.kegg.jp/dbget-bin/www_bget?ds:H00409 ??????

In the first link: I couldn't get the pathways itself, just gene set can be obtained, while in the second link: I can get the pathways itself. So which link do you think I can use?

Answer 2

[xanderpico]

The rWikiPathways package in bioconductor lets you search the database in a few different ways. Like the comments above, exactly how you define a "diabetes pathway" is up to you. For example, you can search for any pathway that mentions "diabetes" anywhere in the title or description and then filter for the human set:

library(rWikiPathways)
db.pathways <- findPathwaysByText('diabetes')  
human.db.pathways <- db.pathways %>% 
    dplyr::filter(species == "Homo sapiens")
human.db.pathways$name

There were 4 that popped up when I ran this snippet:

[1] "Type II diabetes mellitus"

[2] "Insulin signalling in human adipocytes (diabetic condition)"

[3] "Polyol Pathway"

[4] "16p11.2 distal deletion syndrome"

You can retrieve the WPID identifiers and other info from the same returned data frame.

Alternatively, you can seach by your favorite diabetes gene, e.g., INSR:

findPathwaysByXref('INSR','H') # H denotes HGNC identifer for query xref

This returns 27 human pathways! The bioconductor package includes additional vignettes for other common use cases.

[mrashad]

Thanks a lot for this fruitful answer

Answer 3

[darklings]

Try the ReactomeGraph4R package: https://github.com/reactome/ReactomeGraph4R

library(ReactomeGraph4R)
login()
#> Is the url 'http://localhost:7474'? (Yes/no/cancel) 
#> Does Neo4J require authentication? (yes/No/cancel)
#> Successfully connected to the local Reactome Graph Database v76!

## fetch all human (DO) diseases that are stored in the Reactome database
all.diseases <- matchObject(schemaClass="Disease")[["databaseObject"]]
head(all.diseases, 2)
#>   schemaClass identifier                                             synonym
#> 1     Disease        162 malignant tumor, malignant neoplasm, primary cancer
#> 2     Disease      11132                                                NULL
#>   databaseName           displayName    dbId                  name
#> 1         DOID                cancer 1500689                cancer
#> 2         DOID prostatic hypertrophy 9617381 prostatic hypertrophy
#>                                                                                                                                                     definition
#> 1 A disease of cellular proliferation that is malignant and primary, characterized by uncontrolled cellular proliferation, local cell invasion and metastasis.
#> 2                                                                                                                                                         <NA>
#>                                                                 url
#> 1   https://www.ebi.ac.uk/ols/ontologies/doid/terms?obo_id=DOID:162
#> 2 https://www.ebi.ac.uk/ols/ontologies/doid/terms?obo_id=DOID:11132

## select those related to diabetes
diabetes <- all.diseases[grepl("diabetes", all.diseases$displayName),]
diabetes
#>     schemaClass identifier
#> 523     Disease      11717
#> 545     Disease      12388
#> 564     Disease    0050524
#>                                                                                                                        synonym
#> 523                                                                               diabetes mellitus syndrome in newborn infant
#> 545 central diabetes insipidus, Vasopressin deficiency, vasopressin defective diabetes insipidus, Pituitary diabetes insipidus
#> 564                                                                                                  MODY, MASON-TYPE DIABETES
#>     databaseName                          displayName    dbId
#> 523         DOID           neonatal diabetes mellitus 5683190
#> 545         DOID  neurohypophyseal diabetes insipidus 5623603
#> 564         DOID maturity-onset diabetes of the young 5623598
#>                                     name
#> 523           neonatal diabetes mellitus
#> 545  neurohypophyseal diabetes insipidus
#> 564 maturity-onset diabetes of the young
#>                                                                                                  definition
#> 523                                                                                                    <NA>
#> 545                                                                                                    <NA>
#> 564 A genetic disease that has_material_basis_in mutations in the MODY genes disrupting insulin production.
#>                                                                     url
#> 523   https://www.ebi.ac.uk/ols/ontologies/doid/terms?obo_id=DOID:11717
#> 545   https://www.ebi.ac.uk/ols/ontologies/doid/terms?obo_id=DOID:12388
#> 564 https://www.ebi.ac.uk/ols/ontologies/doid/terms?obo_id=DOID:0050524

## fetch all instances associated with those diseases
diabetes.list <- lapply(diabetes$dbId, function(id) {
  matchDiseases(id = id, species = "human", type = "row")
})
#> Retrieving instances associated with the given Disease...
#> Retrieving instances associated with the given Disease...
#> Retrieving instances associated with the given Disease...
str(diabetes.list, max.level = 2)
#> List of 3
#>  $ :List of 3
#>   ..$ disease       :'data.frame':   1 obs. of  8 variables:
#>   ..$ databaseObject:'data.frame':   10 obs. of  18 variables:
#>   ..$ relationships :'data.frame':   10 obs. of  9 variables:
#>  $ :List of 3
#>   ..$ disease       :'data.frame':   1 obs. of  8 variables:
#>   ..$ databaseObject:'data.frame':   12 obs. of  18 variables:
#>   ..$ relationships :'data.frame':   12 obs. of  9 variables:
#>  $ :List of 3
#>   ..$ disease       :'data.frame':   1 obs. of  9 variables:
#>   ..$ databaseObject:'data.frame':   9 obs. of  18 variables:
#>   ..$ relationships :'data.frame':   9 obs. of  9 variables:

## check the 'databaseObject' dataframes
head(diabetes.list[[1]][["databaseObject"]], 2)
#>   schemaClass
#> 1     Pathway
#> 2    Reaction
#>                                                                                                                     displayName
#> 1                                                                           Defective ABCC8 can cause hypo- and hyper-glycemias
#> 2 Activating ABCC8 mutants cause hyperglycemia in permanent neonatal diabetes mellitus (PNDM) and transient neonatal DM (TNDM).
#>      dbId
#> 1 5683177
#> 2 5683209
#>                                                                                                                            name
#> 1                                                                           Defective ABCC8 can cause hypo- and hyper-glycemias
#> 2 Activating ABCC8 mutants cause hyperglycemia in permanent neonatal diabetes mellitus (PNDM) and transient neonatal DM (TNDM).
#>   isInDisease releaseDate          stId  speciesName diagramHeight
#> 1        TRUE  2015-12-15 R-HSA-5683177 Homo sapiens          1915
#> 2        TRUE  2015-12-15 R-HSA-5683209 Homo sapiens            NA
#>       stIdVersion hasDiagram isInferred diagramWidth isChimeric   category
#> 1 R-HSA-5683177.2       TRUE      FALSE         3375         NA       <NA>
#> 2 R-HSA-5683209.1         NA      FALSE           NA      FALSE transition
#>   startCoordinate referenceType endCoordinate
#> 1              NA          <NA>            NA
#> 2              NA          <NA>            NA

^{Created on 2021-04-29 by the [reprex package](https://reprex.tidyverse.org) (v2.0.0)}