I am trying to count the number of paralogues for the mouse homologues of the human protein-coding genes using BioMart. But for example in the 'PLIN4' gene its counting 35,000 paralogues instead of 4.

We think it is because some genes have one to many paralogues which causes repeats. When I run a single gene its gives me back the correct number of paralogues. Is there a way to either remove these repeats from the results or a way around this so that BioMart doesn't output these repeats.

I have also thought of maybe running one gene at a time, then counting it by setting up some sort of loop so that it does all of the genes from the list automatically.

The code I have written so far is:

# Load the biomaRt package:

ensembl_hsapiens <- useMart("ensembl", 
                          dataset = "hsapiens_gene_ensembl")
ensembl_mouse <- useMart("ensembl", 
                       dataset = "mmusculus_gene_ensembl")

# Get all human protein coding genes:

hsapien_PC_genes <- getBM(attributes = c("ensembl_gene_id", 
                          filters = "biotype", 
                          values = "protein_coding", 
                          mart = ensembl_hsapiens)

ensembl_gene_ID <- hsapien_PC_genes$ensembl_gene_id

# Get mouse homologues

mouse_homologues <- getBM(attributes = c("ensembl_gene_id", "external_gene_name", 
                        filters = "ensembl_gene_id", 
                        values = c(ensembl_gene_ID), 
                        mart = ensembl_hsapiens)

# Get mouse external gene name 

mouse_homologues_external_gene_names <- mouse_homologues$mmusculus_homolog_associated_gene_name

mouse_paralogues <- getBM(attributes = c("hsapiens_homolog_associated_gene_name",
                        filters = "external_gene_name", 
                        values = c(mouse_homologues_external_gene_names) , mart = ensembl_mouse)

# Remove genes with no paralogues 
mouse_paralogs_data <- mouse_paralogues[!(is.na(mouse_paralogues$mmusculus_paralog_associated_gene_name)
mouse_paralogues$mmusculus_paralog_associated_gene_name==""), ]

# Count paralogues per gene

count_mouse_paralogues <- count(mouse_paralogs_data, "external_gene_name")

1 Answer 1


I'd use dplyr for this.

First, mouse paralogues contains no values for mmusculus_paralog_associated_gene_name that are NA:

mouse_paralogues %>% 
  filter(is.na(mmusculus_paralog_associated_gene_name)) %>% 

[1] 0

But there are values which are empty strings, so you can filter for those if you wish.

count_mouse_paralogues <- mouse_paralogues %>%
  filter(hsapiens_homolog_associated_gene_name != "") %>%

Confirm that this gives the expected result (n = 4) for Plin4:

count_mouse_paralogues %>%
  filter(external_gene_name == "Plin4")

# A tibble: 1 x 2
  hsapiens_homolog_associated_gene_name     n
  <chr>                                 <int>
1 Plin4                                     4

And check mouse_paralogues:

mouse_paralogues %>% 
  filter(hsapiens_homolog_associated_gene_name == "PLIN4")

  hsapiens_homolog_associated_gene_name external_gene_name mmusculus_paralog_associated_gene_name
1                                 PLIN4              Plin4                                  Plin3
2                                 PLIN4              Plin4                                  Plin5
3                                 PLIN4              Plin4                                  Plin2
4                                 PLIN4              Plin4                                  Plin1

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.