Difference between UCSC and Ensembl genomes

Question

[silas008]

I am really confused about the differences between genomes released by UCSC and Ensembl. What is the difference between them?

For example, mirBase has this miRNA annotation of mice:

Chromosomal coordinates of Mus musculus microRNAs

# microRNAs:               miRBase v21
# genome-build-id:         GRCm38
# genome-build-accession:  NCBI_Assembly:GCA_000001635.2

If I mapped my reads using UCSC mm10 genome can I do the count using this annotation? Or should I do the mapping using Ensembl GRCm38.p4 genome?

Answer 1

[Jean-Karim Heriche]

Both resources use the same assembly but each annotates it using their own pipeline with different gene definitions so you shouldn't expect the annotations to be identical. Whatever you're doing, you should pick one reference on stick to it, a mix and match approach will most of the time give you inconsistent results. You may also find this paper of interest.

[silas008]

Thank you, Jean... great paper. I'm reading it.

Answer 2

[Ian]

My recommendation would be to use UCSC as it is useful for sharing data via "sessions" or "trackhubs". The later can even be used to display data on a genome that is not hosted by UCSC. Be sure to use the european mirror version of UCSC.

[Emily]

You can add track hubs in Ensembl. You can create accounts, save personal data to accounts and share with groups in Ensembl.