Preamble

This document provides the necessary code for creating the rds file with the annotation table used in the course. It uses biomaRt to retrieve the annotation from Ensembl. The list of Ensembl IDs to retrieve annotation for is extracted from an rds object containing the DESeq2 differential expression results, which was already created.

This document is not intended as a tutorial, for a detailed explanation of how to use biomaRt please see supplementary materials Annotation with biomaRt. Nor is this document intended to be a guide for creating your own annotations, some totally arbitrary decisions were made in order to save effort. 

library(DESeq2)
library(biomaRt)
library(tidyverse)

Load a DESeq2 results table to retrieve gene ids from

results.interaction.11 <- readRDS("RObjects/DESeqResults.interaction_d11.rds")

Retrieve annotation using biomaRt

Set up connection to ensembl database

ensembl <- useEnsembl(biomart = 'genes', 
                      dataset = 'mmusculus_gene_ensembl',
                      version = 102)

Retrieve the complete annotation

filterType <- "ensembl_gene_id"
filterValues <- rownames(results.interaction.11)
attributeNames <- c("ensembl_gene_id",
                    "entrezgene_id",
                    "external_gene_name",
                    "description",
                    "gene_biotype",
                    "chromosome_name",
                    "start_position",
                    "end_position",
                    "strand",
                    "entrezgene_accession")

annot <- getBM(attributes=attributeNames,
               filters = filterType,
               values = filterValues,
               mart = ensembl)

This is the complete annotation with all the one-to-many relationships. Export it, so that we won’t have to run the query again if we wish to come back to it.

saveRDS(annot, file="Full_annotation_with_duplicates.rds")

Fix the various One-to-Many relationships

Check for duplicated Entrez IDs

annotUn %>% 
    filter(!is.na(entrezgene_id)) %>% 
    add_count(entrezgene_id) %>% 
    filter(n>1) %>% 
    count(entrezgene_id)
##    entrezgene_id n
## 1          11877 2
## 2          14356 2
## 3          15953 2
## 4          18752 2
## 5          18861 2
## 6          19243 2
## 7          26912 2
## 8          27366 2
## 9          50518 2
## 10         66461 2
## 11         66990 2
## 12         67118 2
## 13         67149 2
## 14         67238 2
## 15         68646 2
## 16         70579 2
## 17         74477 2
## 18         78797 2
## 19         94089 2
## 20         99889 2
## 21        102115 2
## 22        207728 2
## 23        218734 2
## 24        280287 2
## 25        637515 3
## 26        664987 2
## 27        677884 3
## 28     100039796 2
## 29     100041194 2
## 30     100042100 3
## 31     100043914 2

There are 31 Entrez IDs that match multiple Ensembl IDs. Resolving which of these is correct would require more extensive research on the data bases. We need a pragmatic solution for the course. Let’s have alook at the signficance of these genes.

A pragmatic solution for the Entrez IDs

as.data.frame(results.interaction.11) %>% 
    rownames_to_column("ensembl_gene_id") %>% 
    left_join(annotUn, by = "ensembl_gene_id") %>% 
    filter(!is.na(entrezgene_id)) %>% 
    add_count(entrezgene_id) %>% 
    filter(n>1) %>% 
    select(ensembl_gene_id, entrezgene_id, padj, chromosome_name) %>% 
    arrange(padj)
##       ensembl_gene_id entrezgene_id         padj          chromosome_name
## 1  ENSMUSG00000115886     100039796 2.188768e-77 CHR_WSB_EIJ_MMCHR11_CTG1
## 2  ENSMUSG00000078920         15953 3.437499e-60                       11
## 3  ENSMUSG00000078921     100039796 2.667329e-34                       11
## 4  ENSMUSG00000096488     100042100 5.880162e-02                       14
## 5  ENSMUSG00000022253         68646 9.627923e-02                       15
## 6  ENSMUSG00000072812     100041194 9.939717e-02                       12
## 7  ENSMUSG00000116158        280287 1.922501e-01                        1
## 8  ENSMUSG00000070390        637515 1.962265e-01                       11
## 9  ENSMUSG00000089945        677884 2.536130e-01                        4
## 10 ENSMUSG00000116275         70579 3.226737e-01                        1
## 11 ENSMUSG00000110195        207728 3.388073e-01                        7
## 12 ENSMUSG00000118332         67238 3.523385e-01                        5
## 13 ENSMUSG00000074513         99889 3.679077e-01                        3
## 14 ENSMUSG00000111527         68646 3.795302e-01         CHR_MG4288_PATCH
## 15 ENSMUSG00000078532         67149 3.894284e-01                        4
## 16 ENSMUSG00000096793     100042100 4.522771e-01                       14
## 17 ENSMUSG00000102976         70579 5.774588e-01                        1
## 18 ENSMUSG00000116408         94089 6.618999e-01 CHR_WSB_EIJ_MMCHR11_CTG1
## 19 ENSMUSG00000063235         66461 6.767142e-01                        2
## 20 ENSMUSG00000083012         67238 7.066279e-01                        5
## 21 ENSMUSG00000000325         11877 7.075947e-01                       16
## 22 ENSMUSG00000022684         67118 7.175289e-01                       16
## 23 ENSMUSG00000106964         67149 7.280606e-01         CHR_MG4266_PATCH
## 24 ENSMUSG00000078816         18752 7.361723e-01                        7
## 25 ENSMUSG00000079737         67118 7.361723e-01                       16
## 26 ENSMUSG00000107877         74477 7.381156e-01                       11
## 27 ENSMUSG00000114378        218734 7.461406e-01                       14
## 28 ENSMUSG00000117310         19243 7.567281e-01                        1
## 29 ENSMUSG00000075569         18861 7.639096e-01                        5
## 30 ENSMUSG00000079109         18861 7.680020e-01                        5
## 31 ENSMUSG00000024571         27366 7.924897e-01                       18
## 32 ENSMUSG00000116254        637515 8.323231e-01 CHR_CAST_EI_MMCHR11_CTG4
## 33 ENSMUSG00000084897         50518 8.461295e-01                        2
## 34 ENSMUSG00000089847         14356 8.554819e-01                        7
## 35 ENSMUSG00000078905        664987 8.595182e-01                        2
## 36 ENSMUSG00000115074         78797 8.648626e-01                        2
## 37 ENSMUSG00000057130         27366 8.678565e-01                       18
## 38 ENSMUSG00000110860         66461 8.826651e-01          CHR_MG191_PATCH
## 39 ENSMUSG00000114004        102115 8.858318e-01                       10
## 40 ENSMUSG00000006378         26912 8.969835e-01                       15
## 41 ENSMUSG00000006471         78797 8.969835e-01                        2
## 42 ENSMUSG00000026064         19243 9.024507e-01                        1
## 43 ENSMUSG00000097394         66990 9.087617e-01          CHR_MG153_PATCH
## 44 ENSMUSG00000090053        677884 9.161891e-01                        4
## 45 ENSMUSG00000078898     100043914 9.177512e-01                        2
## 46 ENSMUSG00000090691     100042100 9.327098e-01                       14
## 47 ENSMUSG00000033111        218734 9.334572e-01                       14
## 48 ENSMUSG00000097449         18752 9.357120e-01         CHR_MG4151_PATCH
## 49 ENSMUSG00000040350         94089 9.519887e-01                       11
## 50 ENSMUSG00000116378         26912 9.537884e-01                       15
## 51 ENSMUSG00000110234         14356 9.544490e-01                        7
## 52 ENSMUSG00000102805         99889 9.688617e-01                        3
## 53 ENSMUSG00000078440        102115 9.729195e-01                       10
## 54 ENSMUSG00000020807         74477 9.754994e-01                       11
## 55 ENSMUSG00000098615         11877 9.792462e-01  CHR_MG3833_MG4220_PATCH
## 56 ENSMUSG00000024845         66990 9.818681e-01                       19
## 57 ENSMUSG00000030653        207728 9.829405e-01                        7
## 58 ENSMUSG00000090093        664987 9.883529e-01                        2
## 59 ENSMUSG00000038729        677884 9.908806e-01                        4
## 60 ENSMUSG00000115958        280287 9.960817e-01                        1
## 61 ENSMUSG00000027596         50518           NA                        2
## 62 ENSMUSG00000095545     100043914           NA                        2
## 63 ENSMUSG00000116073        637515           NA CHR_PWK_PHJ_MMCHR11_CTG2
## 64 ENSMUSG00000116330         15953           NA CHR_WSB_EIJ_MMCHR11_CTG1
## 65 ENSMUSG00000116477     100041194           NA         CHR_MG3490_PATCH

These genes are mostly non-significant, so it’s really not going to affect the results of the downstream analyses. Seeing as this is only a teaching exercise, we’ll arbitrarily set the second entry to NA. Some of the duplicates are on patch scaffolds, we’ll arrange by chromosome, so that these get set to NA.

dupEntrez <- annotUn %>% 
    add_count(entrezgene_id) %>% 
    filter(n>1) %>% 
    select(-n) %>% 
    arrange(entrezgene_id, chromosome_name)
dupEntrez$entrezgene_id[duplicated(dupEntrez$entrezgene_id)] <- NA

annotFinal <- annotUn %>% 
    add_count(entrezgene_id) %>% 
    filter(n==1) %>% 
    select(-n) %>% 
    bind_rows(dupEntrez)

Final checks

dim(annotFinal)
## [1] 20091    10
annotFinal %>% 
    filter(!is.na(entrezgene_id)) %>% 
    add_count(entrezgene_id) %>% 
    filter(n>1) 
##  [1] ensembl_gene_id      entrezgene_id        external_gene_name  
##  [4] description          gene_biotype         chromosome_name     
##  [7] start_position       end_position         strand              
## [10] entrezgene_accession n                   
## <0 rows> (or 0-length row.names)
all(filterValues%in%annotFinal$ensembl_gene_id)
## [1] TRUE

Final table

ensemblAnnot <- rownames(results.interaction.11) %>%  
    enframe(name = NULL, value = "ensembl_gene_id")  %>%  
    left_join(annotFinal) %>%
    dplyr::select(GeneID="ensembl_gene_id", Entrez="entrezgene_id",
                  Symbol="external_gene_name", Description="description",
                  Biotype="gene_biotype", Chr="chromosome_name",
                  Start="start_position", End="end_position",
                  Strand="strand")

saveRDS(ensemblAnnot, file="RObjects/Ensembl_annotations.rds")