Description
Recount3 is a comprehensive resource for re-analyzing RNA-seq data. It provides uniformly processed
RNA-seq data and associated metadata from a wide range of studies, enabling researchers to access
and analyze gene expression data in a consistent manner. Recount3 aggregates data from multiple
sources, including the
Sequence Read Archive (SRA)
and the
Genotype-Tissue Expression (GTEx) project,
and reprocesses it using a standardized pipeline. This allows for cross-study comparisons and
meta-analyses, facilitating discoveries in genomics and transcriptomics. Processed recount3 data
were integrated into the
Snaptron system
for indexing and querying data summaries. Recount3 is available
at: http://rna.recount.bio.
These tracks display the recount3 intron data, including split read counts and splice junction
motifs. For hg38, tracks are available for GTEx, TCGA, SRA, and CCLE data sources, while mm10
includes the SRA track only.
Display Conventions
Intron items are colored based on splice junction motifs and read support. Darker colors indicate
higher read coverage. Split read counts and splice motifs are shown on mouseover.
By default, only introns with a minimum read count of 10,000 are shown. This threshold can be
changed on the track configuration page.
The intron items are color-coded (darker colors indicate higher coverage):
- Sky blue: GT donors and AG acceptors (CT and AC on
the minus strand)
- Turquoise: GC donors and AG acceptors (CT and GC on the minus strand)
- Orange: AT donors and AC acceptors (GT and AT on the
minus strand)
- Grey: Non-canonical junction motifs. These could be
sequencing errors, polymorphisms, or very rare U12 introns.
Introns can be filtered by:
- Intron size - Length of the intron. The default range is 30 to 100,000 bases.
- Split read count - Number of split reads supporting the intron. The default is a
minimum of 10,000 reads.
- Splice junction motif - The motif is specified in the form GT/AG, with
canonical motifs in uppercase and unknown motifs in lowercase.
The default is no filtering.
- Strand - Filter by positive strand ('+'),
negative strand ('-'), and/or
unknown strand ('.'). The default is no strand filtering ('all').
Methods
A distributed processing system for RNA-seq data called Monorail was developed. Using Monorail,
recount3 processed and summarized 316,443 human and 416,803 mouse RNA-seq run accessions collected
from the Sequence Read Archive (SRA), with the human runs including large-scale consortia such as
GTEx v8 and The Cancer Genome Atlas (TCGA).
Junction files were converted to BED format. For grayscaling total read count was log10
transformed and multiplied by 10 to get a score between 0 and 225, which can be found
in the BED score field.
Data Access
The raw data can be explored interactively with the
Table Browser or the
Data Integrator.
For automated analysis, the data may be queried from our
REST API.
Please refer to our
mailing list archives
for questions or our
Data Access FAQ
for more information.
The original junction files for human can be found at:
The mouse junction file is available at:
References
Wilks C, Zheng SC, Chen FY, Charles R, Solomon B, Ling JP, Imada EL, Zhang D, Joseph L, Leek JT
et al.
recount3: summaries and queries for large-scale RNA-seq expression and splicing.
Genome Biol. 2021 Nov 29;22(1):323.
PMID: 34844637; PMC: PMC8628444