Galaxy |

RNA STAR (version 2.7.5a)

Single-end or paired-end reads:

RNA-Seq FASTQ/FASTA file:

Custom or built-in reference genome:

Built-ins were indexed using default options

Reference genome with or without an annotation:

Select the '... with builtin gene-model' option to select from the list of available indexes that were built with splice junction information. Select the '... without builtin gene-model' option to select from the list of available indexes without annotated splice junctions.

Select reference genome:

If your genome of interest is not listed, contact the Galaxy team

Gene model (gff3,gtf) file for splice junctions:

Exon junction information for mapping splices

Length of the genomic sequence around annotated junctions:

Used in constructing the splice junctions database. Ideal value is ReadLength-1

Use 2-pass mapping for more sensitive novel splice junction discovery:

For a study with multiple samples, multisample 2-pass mapping is the most sensitive approach. It involves two separate runs of STAR for each sample, where, in the second run of each sample, the splice junctions found in any sample in the first runs are treated as additional known junctions. If you plan to use the mapping results as input for STAR-Fusion it is recommended that you use at least single-sample 2-pass mapping of all reads.

twopass_read_subset:

sj_precalculated:

Per gene/transcript output:

STAR can provide analysis results not only with respect to the reference genome, but also with respect to genes and transcripts described by a gene model. Note: This functionality requires either the selection above of a cached index with a gene model, or a gene model provided alongside the index/reference genome in GTF or GFF3 format!

Report chimeric alignments?:

Choose if and how chimeric alignments should be reported. STAR-Fusion users should select the 'Junctions' option and use the resulting tabular dataset as input to STAR-Fusion. Everyone else: note that selecting 'WithinBAM' or 'WithinBAM Junctions' disables the --chimMultimapNmax setting in the algorithmic parameters section below (the tool will only consider uniquely mapped reads in the search for chimeric alignments). If you disable the reporting of chimeric alignments here, then all chimeric alignment settings in the algorithmic parameters section below will be ignored.

BAM output format specifications

BAM output format specification 0

Output filter criterias

Output filter criteria 0

Algorithmic settings

Algorithmic settings 0

Performance tweaks / Troubleshootings

Performance tweaks / Troubleshooting 0

What it does

STAR is an ultrafast universal RNA-seq aligner.

Compatibility Notes

STAR has a huge amount of options to filter alignments and to configure the exact format of its output.

Some tools you may plan to use in your downstream analysis of the results are known to be sensitive to these settings or combinations of them.

STAR-Fusion

STAR-Fusion can use the chimeric junctions output of STAR as input, but you need to enable chimeric alignment detection by STAR for that dataset to be generated. Hence, be sure to select:

Report chimeric alignments?: As separate tabular "Junctions" output (Junctions).

In addition, for best results it is recommended that you

use 2-pass mapping for more sensitive novel splice junction discovery
under BAM output format specification, Read alignment tags to include in the BAM output: select XS as an additional tag to generate (this is the equivalent of using --outSAMstrandField intronMotif on the command line)
under Algorithmic settings, Configure seed, alignment and limits options: use parameters suggested for STAR-Fusion.

Cufflinks

Cufflinks is not considered to be the best tool for use downstream of STAR anymore. Consider using Stringtie instead, which also should pose no compatibility issues.

To avoid compatibility issues with Cufflinks you should:

select XS as a Read alignment tag to include in the BAM output if (and only if) your sequenced reads come from an unstranded library prep
not select the jM and jI tags for inclusion
keep the HI tag selected and
select HI tag values should be zero-based
exclude All alignments across non-canonical junctions under Output filter criteria -> Exclude the following records from the BAM output

Attribution

Minor tweaks to output names to suit downstream purposes, toolshed automated dependencies and odds and ends of other code and documentation comprising this tool were originally written by Ross Lazarus and have been licensed under the creative commons BY-NC_ND 3.0 license.