Dreaming time

Download processed RNA-Seq data from EBI

Ji Huang 2018-02-09 3 min read

Update 2022-09-01 The RNASeq-er API no longer works for me. The FTP site may still work.

For my gene network projects, a majority of my time was spent on manually curate RNA-Seq data on SRA, download SRA data and align them on the maize genome. This week I found an API from EMBL, RNASeq-er API that already aligned RNA-Seq reads on genome for you.

The paper (The RNASeq-er API—a gateway to systematically updated analysis of public RNA-seq data was published on Bioinformatics, 2017.

This post is to show how to do some basic search and download using the API. I followed steps on API’s document.

1. What is RNASeq-er?

RNASeq-er is an API that allow users to download processed RNA-Seq files, including meta-data, cram and processed gene experssion data. It uses iRAP pipeline. It uses TopHat2 and HTSeq2/DEXSeq based on Ensemble genome assembly. See below flowchat.

image

2. Usage exmples

Here I will use maize RNA-Seq study SRP047420 as an example. This study was published by Gent et al 2014 Plant Cell paper. It contains 6 RNA-Seq libraries, mop1-wt vs mop1-mut. Based on the information, all sequences were aligned to AGPv4.

2.1 How many plant species in the database?

You can run wget or curl in Linux console to download the table, or just use plain http and search in Chrome as I did.

https://www.ebi.ac.uk/fg/rnaseq/api/tsv/0/getOrganisms/plants

The result is like below, you will get a tsv table with a) Taxomic name of the organism and b) name of the organism used as reference genome. The /0 in the url is the minimum percentage of reads mapped to the reference genome.

Imgur

2.2 Sample meta-data for one study.

You can get the meta-data for one study by searchinng: https://www.ebi.ac.uk/fg/rnaseq/api/tsv/getSampleAttributesPerRunByStudy/SRP047420

But, I found it’s not very easy to extract useful information. Maybe JSON format will be better.

Imgur

2.3 Alignment files for one study

wget https://www.ebi.ac.uk/fg/rnaseq/api/tsv/70/getRunsByStudy/SRP047420

You will get Sample IDs, Mapping quality, genome version and download files FTP address (including cram, bedgraph and bigwig format).

Imgur

Actually, if you go to the FTP site of one library, for example SRR1583909, there are all files downloadable. These files include gene/transcript level expression, exon level expression, kallisto result and quality control result.

Imgur

2.4 Alignment files for one organism

You can get FTP addresses for all aligned files for one organism (at 70% aligned on genome).

wget https://www.ebi.ac.uk/fg/rnaseq/api/tsv/70/getRunsByOrganism/zea_mays

On 2018-02-09, there all 7147 libraries in maize that have been processed. The last line did not have LF, so if use wc -l, there was one line less.

Imgur

2.5 Gene expression data for one study.

To get gene expresion data for one study:

Search: https://www.ebi.ac.uk/fg/rnaseq/api/tsv/getStudy/SRP047420

Imgur

You can also find other types of gene expression data from the FTP site (ftp://ftp.ebi.ac.uk/pub/databases/arrayexpress/data/atlas/rnaseq/studies/ena/SRP047420/zea_mays)

Imgur

p

2.6 Gene expression data for one organism.

You can also get all gene expression files for one specific organism, for example zea_mays. I did not see kallisto result in the table, but you can download based on the FTP from 2.2.

wget https://www.ebi.ac.uk/fg/rnaseq/api/tsv/getStudiesByOrganism/zea_mays

In total there are 661 studies in maize.

Imgur

2.7 Mapping quality for all organisms

Search: https://www.ebi.ac.uk/fg/rnaseq/api/tsv/getOrganismsMappingQuality

3. Summary

I think this is awesome that EBI already processed the gene expression data FOR YOU! The only thing that can improve is to use HISAT2/STAR instead of TopHat2, since TopHat is out-dated.