STAR manual 2.7.

10b
Alexander Dobin
[email protected]
October 20, 2022

Contents
1 Getting started. 4
1.1 Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.1 Installation - in depth and troubleshooting. . . . . . . . . . . . . . . . . . . . 4
1.2 Basic workflow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2 Generating genome indexes. 5

2.1 Basic options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Advanced options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 Which chromosomes/scaffolds/patches to include? . . . . . . . . . . . . . . . 6
2.2.2 Which annotations to use? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.3 Annotations in GFF format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.4 Using a list of annotated junctions. . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.5 Very small genome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2.6 Genome with a large number of references. . . . . . . . . . . . . . . . . . . . . 7

3 Running mapping jobs. 7

3.1 Basic options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 Mapping multiple files in one run. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.3 Advanced options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.3.1 Using annotations at the mapping stage. . . . . . . . . . . . . . . . . . . . . . 8
3.3.2 ENCODE options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.4 Using shared memory for the genome indexes. . . . . . . . . . . . . . . . . . . . . . 9

4 Output filtering. 10
4.1 Multimappers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

5 Output files. 10
5.1 Log files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.2 SAM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.2.1 Multimappers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1
 5.2.2 SAM attributes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.2.3 Compatibility with Cufflinks/Cuffdiff. . . . . . . . . . . . . . . . . . . . . . . 13
5.3 Unsorted and sorted-by-coordinate BAM. . . . . . . . . . . . . . . . . . . . . . . . . 13
5.4 Unmapped reads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.5 Splice junctions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

6 Chimeric and circular alignments. 15

6.1 STAR-Fusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2 Chimeric alignments in the main BAM files. . . . . . . . . . . . . . . . . . . . . . . 15
6.3 Chimeric alignments in Chimeric.out.sam . . . . . . . . . . . . . . . . . . . . . . . 16
6.4 Chimeric alignments in Chimeric.out.junction . . . . . . . . . . . . . . . . . . . . 16

7 Output in transcript coordinates. 18

8 Counting number of reads per gene. 18

9 2-pass mapping. 19
9.1 Multi-sample 2-pass mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
9.2 Per-sample 2-pass mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
9.3 2-pass mapping with re-generated genome. . . . . . . . . . . . . . . . . . . . . . . . 19

10 Merging and mapping of overlapping paired-end reads. 20

11 Detection of personal variants overlapping alignments. 20

12 WASP filtering of allele specific alignments. 20

13 STARconsensus 21

14 Detection of multimapping chimeras. 21

15 STARsolo: mapping, demultiplexing and gene quantification for single cell

RNA-seq 21
15.1 Feature statistics summaries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

16 Description of all options. 23

16.1 Parameter Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
16.2 System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
16.3 Run Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
16.4 Genome Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
16.5 Genome Indexing Parameters - only used with –runMode genomeGenerate . . . . . 26
16.6 Splice Junctions Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
16.7 Variation parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
16.8 Input Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
16.9 Read Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
16.10 Read Clipping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
16.11 Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

2
16.12 Output: general . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
16.13 Output: SAM and BAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
16.14 BAM processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
16.15 Output Wiggle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
16.16 Output Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
16.17 Output splice junctions (SJ.out.tab) . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
16.18 Output Filtering: Splice Junctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
16.19 Scoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
16.20 Alignments and Seeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
16.21 Paired-End reads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
16.22 Windows, Anchors, Binning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
16.23 Chimeric Alignments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
16.24 Quantification of Annotations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
16.25 2-pass Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
16.26 WASP parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
16.27 STARsolo (single cell RNA-seq) parameters . . . . . . . . . . . . . . . . . . . . . . . 55

3
1 Getting started.
1.1 Installation.
STAR source code and binaries can be downloaded from GitHub: named releases from https:
//github.com/alexdobin/STAR/releases, or the master branch from https://github.com/
alexdobin/STAR. The pre-compiled STAR executables are located bin/ subdirectory. The static
executables are the easisest to use, as they are statically compiled and are not dependents on external
libraries.
To compile STAR from sources run make in the source directory for a Linux-like environment,
or run make STARforMac for Mac OS X. This will produce the executable ’STAR’ inside the source
directory.

1.1.1 Installation - in depth and troubleshooting.

STAR is compiled with gcc c++ compiler and depends only on standard gcc libraries. Some generic
instructions on installing correct gcc environments are given below.

Ubuntu.
$ sudo apt-get update
$ sudo apt-get install g++
$ sudo apt-get install make

Red Hat, CentOS, Fedora.

$ sudo yum update
$ sudo yum install make
$ sudo yum install gcc-c++
$ sudo yum install glibc-static

SUSE.
$ sudo zypper update
$ sudo zypper in gcc gcc-c++

Mac OS X.
Current versions of Mac OS X Xcode are shipped with Clang replacing the standard gcc compiler.
Presently, standard Clang does not support OpenMP which creates problems for STAR compilation.
One option to avoid this problem is to install gcc (preferably using homebrew package manager).
Another option is to add OpenMP functionality to Clang.

1.2 Basic workflow.

Basic STAR workflow consists of 2 steps:
1. Generating genome indexes files (see Section 2. Generating genome indexes.
In this step user supplied the reference genome sequences (FASTA files) and annota-
tions (GTF file), from which STAR generate genome indexes that are utilized in the

4
 2nd (mapping) step. The genome indexes are saved to disk and need only be generated
once for each genome/annotation combination. A limited collection of STAR genomes
is available from http://labshare.cshl.edu/shares/gingeraslab/www-data/dobin/STAR/
STARgenomes/, however, it is strongly recommended that users generate their own genome
indexes with most up-to-date assemblies and annotations.
2. Mapping reads to the genome (see Section 3. Running mapping jobs).
In this step user supplies the genome files generated in the 1st step, as well as the RNA-seq
reads (sequences) in the form of FASTA or FASTQ files. STAR maps the reads to the genome,
and writes several output files, such as alignments (SAM/BAM), mapping summary statistics,
splice junctions, unmapped reads, signal (wiggle) tracks etc. Output files are described in
Section 5. Output files. Mapping is controlled by a variety of input parameters (options) that
are described in brief in Section 3. Running mapping jobs, and in more detail in Section 16.
Description of all options.
STAR command line has the following format:
STAR --option1-name option1-value(s)--option2-name option2-value(s) ...
If an option can accept multiple values, they are separated by spaces, and in a few cases - by commas.

2 Generating genome indexes.

2.1 Basic options.
The basic options to generate genome indices are as follows:
--runThreadN NumberOfThreads
--runMode genomeGenerate
--genomeDir /path/to/genomeDir
--genomeFastaFiles /path/to/genome/fasta1 /path/to/genome/fasta2 ...
--sjdbGTFfile /path/to/annotations.gtf
--sjdbOverhang ReadLength-1
--runThreadN option defines the number of threads to be used for genome generation, it has
to be set to the number of available cores on the server node.
--runMode genomeGenerate option directs STAR to run genome indices generation job.
--genomeDir specifies path to the directory (henceforth called ”genome directory” where the
genome indices are stored. This directory has to be created (with mkdir) before STAR run
and needs to have writing permissions. The file system needs to have at least 100GB of disk
space available for a typical mammalian genome. It is recommended to remove all files from
the genome directory before running the genome generation step. This directory path will have
to be supplied at the mapping step to identify the reference genome.
--genomeFastaFiles specifies one or more FASTA files with the genome reference sequences.
Multiple reference sequences (henceforth called “chromosomes”) are allowed for each fasta file.
You can rename the chromosomes’ names in the chrName.txt keeping the order of the chromo-
somes in the file: the names from this file will be used in all output alignment files (such as
.sam). The tabs are not allowed in chromosomes’ names, and spaces are not recommended.

5
 --sjdbGTFfile specifies the path to the file with annotated transcripts in the standard GTF
format. STAR will extract splice junctions from this file and use them to greatly improve
accuracy of the mapping. While this is optional, and STAR can be run without annotations,
using annotations is highly recommended whenever they are available. Starting from 2.4.1a,
the annotations can also be included on the fly at the mapping step.
--sjdbOverhang specifies the length of the genomic sequence around the annotated junction
to be used in constructing the splice junctions database. Ideally, this length should be equal
to the ReadLength-1, where ReadLength is the length of the reads. For instance, for Illumina
2x100b paired-end reads, the ideal value is 100-1=99. In case of reads of varying length, the
ideal value is max(ReadLength)-1. In most cases, the default value of 100 will work as
well as the ideal value.
Genome files comprise binary genome sequence, suffix arrays, text chromosome names/lengths,
splice junctions coordinates, and transcripts/genes information. Most of these files use internal
STAR format and are not intended to be utilized by the end user. It is strongly not recommended
to change any of these files with one exception: you can rename the chromosome names in the
chrName.txt while keeping the order of the chromosomes in this file: the chromosome names from
this file will be used in all output files (e.g. SAM/BAM).

2.2 Advanced options.

2.2.1 Which chromosomes/scaffolds/patches to include?
It is strongly recommended to include major chromosomes (e.g., for human chr1-22,chrX,chrY,chrM,)
as well as un-placed and un-localized scaffolds. Typically, un-placed/un-localized scaffolds add just a
few MegaBases to the genome length, however, a substantial number of reads may map to ribosomal
RNA (rRNA) repeats on these scaffolds. These reads would be reported as unmapped if the scaffolds
are not included in the genome, or, even worse, may be aligned to wrong loci on the chromosomes.
Generally, patches and alternative haplotypes should not be included in the genome.
Examples of acceptable genome sequence files:
ˆ ENSEMBL: files marked with .dna.primary.assembly, such as: ftp://ftp.ensembl.
org/pub/release-77/fasta/homo_sapiens/dna/Homo_sapiens.GRCh38.dna.primary_
assembly.fa.gz
ˆ GENCODE: files marked with PRI (primary). Strongly recommended for mouse and human:
http://www.gencodegenes.org/.

2.2.2 Which annotations to use?

The use of the most comprehensive annotations for a given species is strongly recommended. Very
importantly, chromosome names in the annotations GTF file have to match chromosome names in the
FASTA genome sequence files. For example, one can use ENSEMBL FASTA files with ENSEMBL
GTF files, and UCSC FASTA files with UCSC FASTA files. However, since UCSC uses chr1, chr2,
... naming convention, and ENSEMBL uses 1, 2, ... naming, the ENSEMBL and UCSC FASTA
and GTF files cannot be mixed together, unless chromosomes are renamed to match between the
FASTA anf GTF files.

6
2.2.3 Annotations in GFF format.
In addition to the aforementioned options, for GFF3 formatted annotations you need to use
--sjdbGTFtagExonParentTranscript Parent. In general, for --sjdbGTFfile files STAR only
processes lines which have --sjdbGTFfeatureExon (=exon by default) in the 3rd field (col-
umn). The exons are assigned to the transcripts using parent-child relationship defined by the
--sjdbGTFtagExonParentTranscript (=transcript id by default) GTF/GFF attribute.

2.2.4 Using a list of annotated junctions.

STAR can also utilize annotations formatted as a list of splice junctions coordinates in a text file:
--sjdbFileChrStartEnd /path/to/sjdbFile.txt. This file should contains 4 columns separated
by tabs:
Chr \tab Start \tab End \tab Strand=+/-/.
Here Start and End are first and last bases of the introns (1-based chromosome coordinates). This
file can be used in addition to the --sjdbGTFfile, in which case STAR will extract junctions from
both files.
Note, that the --sjdbFileChrStartEnd file can contain duplicate (identical) junctions, STAR
will collapse (remove) duplicate junctions.

2.2.5 Very small genome.

For small genomes, the parameter --genomeSAindexNbases must to be scaled down, with a typical
value of min(14, log2(GenomeLength)/2 - 1). For example, for 1 megaBase genome, this is equal
to 9, for 100 kiloBase genome, this is equal to 7.

2.2.6 Genome with a large number of references.

If you are using a genome with a large (>5,000) number of references (chro-
somes/scaffolds), you may need to reduce the --genomeChrBinNbits to reduce RAM
consumption. The following scaling is recommended: --genomeChrBinNbits =
min(18,log2[max(GenomeLength/NumberOfReferences,ReadLength)]). For example, for
3 gigaBase genome with 100,000 chromosomes/scaffolds, this is equal to 15.

3 Running mapping jobs.

3.1 Basic options.
The basic options to run a mapping job are as follows:
--runThreadN NumberOfThreads
--genomeDir /path/to/genomeDir
--readFilesIn /path/to/read1 [/path/to/read2 ]

--genomeDir specifies path to the genome directory where genome indices where generated
(see Section 2. Generating genome indexes).

7
 --readFilesIn name(s) (with path) of the files containing the sequences to be mapped (e.g.
RNA-seq FASTQ files). If using Illumina paired-end reads, the read1 and read2 files have to
be supplied. STAR can process both FASTA and FASTQ files. Multi-line (i.e. sequence split
in multiple lines) FASTA (but not FASTQ) files are supported.
If the read files are compressed, use the --readFilesCommand UncompressionCommand option,
where UncompressionCommand is the un-compression command that takes the file name as
input parameter, and sends the uncompressed output to stdout. For example, for gzipped
files (*.gz) use --readFilesCommand zcat OR --readFilesCommand gunzip -c. For bzip2-
compressed files, use --readFilesCommand bunzip2 -c.

3.2 Mapping multiple files in one run.

Multiple samples can be mapped in one run with a single output. This is equivalent to concatenating
the read files before mapping, except that distinct read groups can be used in --outSAMattrRGline
command to keep track of reads from different files. For single-end reads use a comma separated list
(no spaces around commas), e.g.:
--readFilesIn sample1.fq,sample2.fq,sample3.fq
For paired-end reads, use comma separated list for read1, followed by space, followed by comma
separated list for read2, e.g.:
--readFilesIn s1read1.fq,s2read1.fq,s3read1.fq s1read2.fq,s2read2.fq,s3read2.fq
For multiple read files, the corresponding read groups can be supplied with space/comma/space-
separated list in --outSAMattrRGline, e.g.
--outSAMattrRGline ID:sample1 , ID:sample2 , ID:sample3
Note that this list is separated by commas surrounded by spaces (unlike --readFilesIn list).
Another option for mapping multiple reads files, especially convenient for a very large number of
files, is to create a file manifest and supply it in --readFilesManifest /path/to/manifest.tsv.
The manifest file should contain 3 tab-separated columns. For paired-end reads:
read1-file-name tab read2-file-name tab read-group-line
For single-end reads, the 2nd column should contain the dash -:
read1-file-name tab - tab read-group-line
Spaces, but not tabs are allowed in the file names. If read-group-line does not start with ID:,
it can only contain one ID field, and ID: will be added to it. If read-group-line starts with ID:, it
can contain several fields separated by tab, and all the fields will be copied verbatim into SAM @RG
header line.

3.3 Advanced options.

There are many advanced options that control STAR mapping behavior. All options are briefly
described in the Section Section 16. Description of all options.

3.3.1 Using annotations at the mapping stage.

Since 2.4.1a, the annotations can be included on the fly at the mapping step, without including
them at the genome generation step. You can specify --sjdbGTFfile /path/to/ann.gtf and/or
--sjdbFileChrStartEnd /path/to/sj.tab, as well as --sjdbOverhang, and any other --sjdb*

8
options. The genome indices can be generated with or without another set of annotations/junctions.
In the latter case the new junctions will added to the old ones. STAR will insert the junctions into
genome indices on the fly before mapping, which takes 1 2 minutes. The on the fly genome indices
can be saved (for reuse) with --sjdbInsertSave All, into STARgenome directory inside the current
run directory.

3.3.2 ENCODE options

An example of ENCODE standard options for long RNA-seq pipeline is given below:

--outFilterType BySJout
reduces the number of ”spurious” junctions

--outFilterMultimapNmax 20
max number of multiple alignments allowed for a read: if exceeded, the read is considered
unmapped

--alignSJoverhangMin 8
minimum overhang for unannotated junctions

--alignSJDBoverhangMin 1
minimum overhang for annotated junctions

--outFilterMismatchNmax 999
maximum number of mismatches per pair, large number switches off this filter

--outFilterMismatchNoverReadLmax 0.04
max number of mismatches per pair relative to read length: for 2x100b, max number of mis-
matches is 0.04*200=8 for the paired read

--alignIntronMin 20
minimum intron length

--alignIntronMax 1000000
maximum intron length

--alignMatesGapMax 1000000
maximum genomic distance between mates

3.4 Using shared memory for the genome indexes.

The --genomeLoad option controls how the genome is loaded into memory. By default, --genomeLoad
NoSharedMemory, shared memory is not used.
With --genomeLoad LoadAndKeep, STAR loads the genome as a standard Linux shared memory
piece. The genomes are identified by their unique directory paths. Before loading the genome, STAR
checks if the genome has already been loaded into the shared memory. If the genome has not been
loaded, STAR will load it and will keep it in memory even after STAR job finishes. The genome
will be shared with all the other STAR jobs. You can remove the genome from the shared memory

9
running STAR with --genomeLoad Remove. The shared memory piece will be physically removed
only after all STAR jobs attached to it complete. With --genomeLoad LoadAndRemove, STAR will
load genome in the shared memory, and mark it for removal, so that the genome will be removed
from the shared memory once all STAR jobs using it exit. --genomeLoad LoadAndExit, STAR will
load genome in the shared memory, and immediately exit, keeping the genome loaded in the shared
memory for the future runs.
If you need to check or remove shared memory pieces manually, use the standard Linux command
ipcs and ipcrm. If the genome residing in shared memory is not used for a long time it may get paged
out of RAM which will slow down STAR runs considerably. It is strongly recommended to regularly
re-load (i.e. remove and load again) the shared memory genomes.
Many standard Linux distributions do not allow large enough shared memory blocks. You can fix
this issue if you have root privileges, or ask you system administrator to do it. To enable the shared
memory modify or add the following lines to /etc/sysctl.conf:
kernel.shmmax = Nmax
kernel.shmall = Nall
N max, N all numbers should be chosen as follows:
N max > GenomeIndexSize = Genome + SA + SAindex ( 31000000000 for human genome)
N all > GenomeIndexSize/P ageSize
where PageSize is typically 4096 (this can be checked with getconf PAGE SIZE). Then run:
/sbin/sysctl -p
This will increase the allowed shared memory blocks to 31GB, enough for human or mouse genome.

4 Output filtering.
4.1 Multimappers.
The output of multimappers (i.e. reads mapping to multiple loci) is controlled by
--outFilterMultimapNmax N. By default, N=10. If a read maps to ≤N loci, it will be output,
otherwise, it will be considered as unmapped and reported as ”Multimapping: mapped to too many
loci” in the Log.final.out summary statistics file.
The detection of multimappers is controlled by --winAnchorMultimapNmax option, =50 by
default. This parameter should be set to at least the number of multimapping loci, i.e.
--winAnchorMultimapNmax ≥ --outFilterMultimapNmax. Note that this parameter also controls
the overall sensitivity of mapping: increasing it will change (improve) the mapping of unique mappers
as well, though at the cost of slower speed.

5 Output files.
STAR produces multiple output files. All files have standard name, however, you can change the file
prefixes using --outFileNamePrefix /path/to/output/dir/prefix. By default, this parameter is
./, i.e. all output files are written in the current directory.

10
5.1 Log files.
Log.out: main log file with a lot of detailed information about the run. This file is most useful
for troubleshooting and debugging.

Log.progress.out: reports job progress statistics, such as the number of processed reads, %
of mapped reads etc. It is updated in 1 minute intervals.

Log.final.out: summary mapping statistics after mapping job is complete, very useful for
quality control. The statistics are calculated for each read (single- or paired-end) and then
summed or averaged over all reads. Note that STAR counts a paired-end read as one read,
(unlike the samtools flagstat/idxstats, which count each mate separately). Most of the informa-
tion is collected about the UNIQUE mappers (unlike samtools flagstat/idxstats which does not
separate unique or multi-mappers). Each splicing is counted in the numbers of splices, which
would correspond to summing the counts in SJ.out.tab. The mismatch/indel error rates are
calculated on a per base basis, i.e. as total number of mismatches/indels in all unique mappers
divided by the total number of mapped bases.

5.2 SAM.
Aligned.out.sam - alignments in standard SAM format.

5.2.1 Multimappers.
The number of loci Nmap a read maps to is given by NH:i:Nmap field. Value of 1 corresponds to
unique mappers, while values >1 corresponds to multi-mappers. HI attrbiutes enumerates multiple
alignments of a read starting with 1 (this can be changed with the --outSAMattrIHstart - setting
it to 0 may be required for compatibility with downstream software such as Cufflinks).
The mapping quality MAPQ (column 5) is 255 for uniquely mapping reads, and int(-10*log10(1-
1/Nmap)) for multi-mapping reads. This scheme is same as the one used by TopHat and is com-
patible with Cufflinks. The default MAPQ=255 for the unique mappers maybe changed with
--outSAMmapqUnique parameter (integer 0 to 255) to ensure compatibility with downstream tools
such as GATK.
For multi-mappers, all alignments except one are marked with 0x100 (secondary alignment) in
the FLAG (column 2 of the SAM). The unmarked alignment is selected from the best ones (i.e.
highest scoring). This default behavior can be changed with --outSAMprimaryFlag AllBestScore
option, that will output all alignments with the best score as primary alignments (i.e. 0x100 bit in
the FLAG unset).
By default, the order of the multi-mapping alignments for each read is not truly random. The
--outMultimapperOrder Random option outputs multiple alignments for each read in random order,
and also also randomizes the choice of the primary alignment from the highest scoring alignments.
Parameter --runRNGseed can be used to set the random generator seed. With this option, the
ordering of multi-mapping alignments of each read, and the choice of the primary alignment will vary
from run to run, unless only one thread is used and the seed is kept constant.
The --outSAMmultNmax parameter limits the number of output alignments (SAM lines) for
multimappers. For instance, --outSAMmultNmax 1 will output exactly one SAM line for each

11
mapped read. Note that NH:i: tag in STAR will still report the actual number of loci that
the reads map to, while the the number of reported alignments for a read in the SAM file is
min(NH,--outSAMmultNMax). If --outSAMmultNmax is equal to -1, all the alignments are out-
put according to the order specified in --outMultimapperOrder option. If --outSAMmultNmax
is not equal to -1, than top-scoring alignments will always be output first, even for the default
--outMultimapperOrder Old 2.4 option.

5.2.2 SAM attributes.

The SAM attributes can be specified by the user using --outSAMattributes A1 A2 A3 ... option
which accept a list of 2-character SAM attributes. The attributes can be listed in any order, and will
be recorded in that order in the SAM file. By default, STAR outputs NH HI AS nM attributes.

Presets:
None : No SAM attributes
Standard : NH HI AS nM
All : NH HI AS nM NM MD jM jI MC ch
Alignment:
NH : number of loci the reads maps to: = 1 for unique mappers, > 1 for multimappers. Standard
SAM tag.
HI : multiple alignment index, starts with –outSAMattrIHstart (= 1 by default). Standard
SAM tag.
AS : local alignment score, +1/ − 1 for matches/mismateches, score* penalties for indels and
gaps. For PE reads, total score for two mates. Standard SAM tag.
NM : edit distance to the reference (number of mismatched + inserted + deleted bases) for each
mate. Standard SAM tag.
nM : number of mismatches per (paired) alignment, not to be confused with NM, which is the
number of mismatches+indels in each mate.
jM:B:c,M1,M2,... : intron motifs for all junctions (i.e. N in CIGAR): 0: non-canonical;
1: GT/AG, 2: CT/AC, 3: GC/AG, 4: CT/GC, 5: AT/AC, 6: GT/AT. If splice junctions
database is used, and a junction is annotated, 20 is added to its motif value.
MD : string encoding mismatched and deleted reference bases (see standard SAM specifications).
Standard SAM tag.
jI:B:I,Start1,End1,Start2,End2,... : Start and End of introns for all junctions (1-based).
jM jI : attributes require samtools 0.1.18 or later, and were reported to be incompatible with
some downstream tools such as Cufflinks.
Variation:
vA : variant allele.
vG : genomic coordinate of the variant overlapped by the read.

12
 vW : WASP filtering tag, see detailed description in Section 12. Requires --waspOutputMode
SAMtag.
STARsolo:
CR CY UR UY : sequences and quality scores of cell barcodes and UMIs for the solo* demulti-
plexing, not error corrected.
GX GN : gene ID and name.
CB UB : error-corrected cell barcodes and UMIs for solo* demultiplexing. Requires
--outSAMtype BAM SortedByCoordinate.
sM : assessment of CB and UMI.
sS : sequence of the entire barcode (CB,UMI,adapter...).
sQ : quality of the entire barcode.
Unmapped reads:
uT : for unmapped reads, reason for not mapping:
0 : no acceptable seed/windows, ”Unmapped other” in the Log.final.out
1 : best alignment shorter than min allowed mapped length, ”Unmapped: too short” in
the Log.final.out
2 : best alignment has more mismatches than max allowed number of mismatches, ”Un-
mapped: too many mismatches” in the Log.final.out
3 : read maps to more loci than the max number of multimappng loci, ”Multimapping:
mapped to too many loci” in the Log.final.out
4 : unmapped mate of a mapped paired-end read

5.2.3 Compatibility with Cufflinks/Cuffdiff.

For unstranded RNA-seq data, Cufflinks/Cuffdiff require spliced alignments with XS strand attribute,
which STAR will generate with --outSAMstrandField intronMotif option. As required, the XS
strand attribute will be generated for all alignments that contain splice junctions. The spliced
alignments that have undefined strand (i.e. containing only non-canonical unannotated junctions)
will be suppressed.
If you have stranded RNA-seq data, you do not need to use any specific STAR options. Instead,
you need to run Cufflinks with the library option --library-type options. For example, cufflinks
... --library-type fr-firststrand should be used for the “standard” dUTP protocol, includ-
ing Illumina’s stranded Tru-Seq. This option has to be used only for Cufflinks runs and not for STAR
runs.
In addition, it is recommended to remove the non-canonical junctions for Cufflinks runs using
--outFilterIntronMotifs RemoveNoncanonical.

5.3 Unsorted and sorted-by-coordinate BAM.

STAR can output alignments directly in binary BAM format, thus saving time on converting SAM
files to BAM. It can also sort BAM files by coordinates, which is required by many downstream
applications.

13
 --outSAMtype BAM Unsorted
output unsorted Aligned.out.bam file. The paired ends of an alignment are always adjacent,
and multiple alignments of a read are adjacent as well. This ”unsorted” file can be directly
used with downstream software such as HTseq, without the need of name sorting. The order
of the reads will match that of the input FASTQ(A) files only if one thread is used
--runThread 1, and --outFilterType --BySJout is not used.

--outSAMtype BAM SortedByCoordinate

output sorted by coordinate Aligned.sortedByCoord.out.bam file, similar to samtools
sort command. If this option causes problems, it is recommended to reduce
--outBAMsortingThreadN from the default 6 to lower values (as low as 1).

--outSAMtype BAM Unsorted SortedByCoordinate

output both unsorted and sorted files.

5.4 Unmapped reads.

Unmapped reads can be output into the SAM/BAM Aligned.* file(s) with --outSAMunmapped
Within option. --outSAMunmapped Within KeepPairs will (redundantly) record unmapped mate
for each alignment, and, in case of unsorted output, keep it adjacent to its mapped mate (this only
affects multi-mapping reads). uT SAM tag indicates reason for not mapping:
0 : no acceptable seed/windows, ”Unmapped other” in the Log.final.out
1 : best alignment shorter than min allowed mapped length, ”Unmapped: too short” in the
Log.final.out
2 : best alignment has more mismatches than max allowed number of mismatches, ”Unmapped:
too many mismatches” in the Log.final.out
3 : read maps to more loci than the max number of multimappng loci, ”Multimapping: mapped
to too many loci” in the Log.final.out
4 : unmapped mate of a mapped paired-end read
--outReadsUnmapped Fastx will output unmapped reads into separate file(s)
Unmapped.out.mate1[2], formatted the same way as input read files (i.e. FASTQ or FASTA). For
paired-end reads, if a read maps as a whole, but one of the mates does not map, both mates will
also be output in Unmapped.out.mate1/2 files. To indicate the mapping status of the read mates,
the following tags are appended to the read name:
00: mates were not mapped;
10: 1st mate mapped, 2nd unmapped
01: 1st unmapped, 2nd mapped

5.5 Splice junctions.

SJ.out.tab contains high confidence collapsed splice junctions in tab-delimited format. Note that
STAR defines the junction start/end as intronic bases, while many other software define them as
exonic bases. The columns have the following meaning:
column 1: chromosome

14
 column 2: first base of the intron (1-based)

column 3: last base of the intron (1-based)

column 4: strand (0: undefined, 1: +, 2: -)

column 5: intron motif: 0: non-canonical; 1: GT/AG, 2: CT/AC, 3: GC/AG, 4: CT/GC, 5:

AT/AC, 6: GT/AT

column 6: 0: unannotated, 1: annotated in the splice junctions database. Note that in 2-pass
mode, junctions detected in the 1st pass are reported as annotated, in addition to
annotated junctions from GTF.

column 7: number of uniquely mapping reads crossing the junction

column 8: number of multi-mapping reads crossing the junction

column 9: maximum spliced alignment overhang

The filtering for this output file is controlled by the --outSJfilter* parameters, as described in
Section 16.18. Output Filtering: Splice Junctions.

6 Chimeric and circular alignments.

To switch on detection of chimeric (fusion) alignments (in addition to normal mapping),
--chimSegmentMin should be set to a positive value. Each chimeric alignment consists of two ”seg-
ments”. Each segment is non-chimeric on its own, but the segments are chimeric to each other (i.e.
the segments belong to different chromosomes, or different strands, or are far from each other). Both
segments may contain splice junctions, and one of the segments may contain portions of both mates.
--chimSegmentMin parameter controls the minimum mapped length of the two segments that is
allowed. For example, if you have 2x75 reads and used --chimSegmentMin 20, a chimeric alignment
with 130b on one chromosome and 20b on the other will be output, while 135 + 15 won’t be.

6.1 STAR-Fusion.
STAR-Fusion is a software package for detecting fusion transcript from STAR chimeric output. It
is developed and maintained by Brian Haas (@Broad Institute), whose effort was inspired by earlier
work done by Nicolas Stransky in the landmark publication ”The landscape of kinase fusions in
cancer” by Stransky et al., Nat Commun 2014, in addition to very nice work done by Daniel Nicorici
with his FusionCatcher software. Please visit its GitHub page for instructions and documentation:
https://github.com/STAR-Fusion/STAR-Fusion.

6.2 Chimeric alignments in the main BAM files.

Chimeric alignments can be included together with normal alignments in the main (sorted or un-
sorted) BAM file(s) using --chimOutType WithinBAM. In these files, formatting of chimeric alignments
follows the latest SAM/BAM specifications.

15
6.3 Chimeric alignments in Chimeric.out.sam .
With --chimOutType SeparateSAMold STAR will output normal alignments into
Aligned.*.sam/bam, and will output chimeric alignments into a separate file Chimeric.out.sam.
Note that this option will be deprecated in the future, and the --chimOutType WithinBAM is strongly
recommended. Some reads may be output to both normal SAM/BAM files, and Chimeric.out.sam
for the following reason. STAR will output a non-chimeric alignment into Aligned.out.sam with
soft-clipping a portion of the read. If this portion is long enough, and it maps well and uniquely some-
where else in the genome, there will also be a chimeric alignment output into Chimeric.out.sam.
For instance, if you have a paired-end read where the second mate can be split chimerically into
70 and 30 bases. The 100b of the first mate + 70b of the 2nd mate map non-chimerically,and
the mapping length/score are big enough, so they will be output into Aligned.out.sam file. At
the same time, the chimeric segments 100-mate1 + 70-mate2 and 30-mate2 will be output into
Chimeric.out.sam.

6.4 Chimeric alignments in Chimeric.out.junction

By default, or with --chimOutType Junctions, STAR will generate Chimeric.out.junction file
which maybe more convenient for downstream analysis. The format of this file is as follows. Every
line contains one chimerically aligned read, e.g.:

chr22 23632601 + chr9 133729450 + 1 0 0

SINATRA-0006:3:3:6387:5665#0 23632554 47M29S 133729451 47S29M40p76M

The first 9 columns give information about the chimeric junction:

column 1: chr donorA : chromosome of the donor

column 2: brkpt donorA : first base of the intron of the donor (1-based)

column 3: strand donorA : strand of the donor

column 4: chr acceptorB : chromosome of the acceptor

column 5: brkpt acceptorB : first base of the intron of the acceptor (1-based)

column 6: strand acceptorB : strand of the acceptor

column 7: junction type : -1=encompassing junction (between the mates), 1=GT/AG,

2=CT/AC

column 8: repeat left lenA : repeat length to the left of the junction

column 9: repeat right lenB : repeat length to the right of the junction

Columns 10-14 describe the alignments of the two chimeric segments, it is SAM like. Alignments
are given with respect to the (+) strand

column 10: read name : name of the RNA-seq fragment

16
 column 11: start alnA : first base of the first segment (on the + strand)

column 12: cigar alnA : CIGAR of the first segment

column 13: start alnB : first base of the second segment

column 14: cigar alnB : CIGAR of the second segment

Columns 15-20 provide alignment score information and relevant metadata. These columns are
only output for multimapping chimeriuc algorithm --chimMultimapNmax >0.

column 15: num chim aln : number of sufficiently scoring chimeric alignments reported for this
RNA-seq fragment.

column 16: max poss aln score : maximum possible alignment score for this fragment’s read(s).

column 17: non chim aln score : best non-chimeric alignment score

column 18: this chim aln score : score for this individual chimeric alignment

column 19: bestall chim aln score : the highest chimeric alignment score encountered for this
RNA-seq fragment among the num chim aln reported chimeric alignments.

column 20: PEmerged bool : boolean indicating that overlapping PE reads were first merged
into a single contiguous sequence before alignment.

column 21: readgrp : read group assignment for the read as indicated in the BAM file

Unlike standard SAM, both mates are recorded in one line here. The gap of length L between the
mates is marked by the p in the CIGAR string. If the mates overlap, L<0.
For strand definitions, when aligning paired end reads, the sequence of the second mate is reverse
complemented.
For encompassing junctions, i.e. junction type: -1=junction is between the mates, columns 2 and
5 represent the bounds on the chimeric junction loci. For the 1st mate, it will be the genomic base
following the last 3’ mapped base. For the 2nd mate (which is reverse complemented to have the
same orientation as 1st mate), it will be the genomic base preceding the 5’ mapped base. For exam-
ple, if there is a chimeric junction that connects chr1/+strand/base1000 to chr2/+strand/base2000,
and read 1 maps to chr1/+strand/bases800-900, and read 2 (after reverse complementing) maps to
chr2/+strand/bases2100-2200, then columns 2 and 5 will have 901 and 2099.
To filter chimeric junctions and find the number of reads supporting each junction you could use,
for example:

cat Chimeric.out.junction |
awk ’$1!="chrM" && $4!="chrM" && $7>0 && $8+$9<=5 {print $1,$2,$3,$4,$5,$6,$7,$8,$9}’ |
sort | uniq -c | sort -k1,1rn

This will keep only the canonical junctions with the repeat length less than 5 and will remove chimeras
with mitochondrion genome.
When I do it for one of our K562 runs, I get:

17
181 chr1 144676873 - chr1 147917466 + 1 0 1
29 chr5 69515744 - chr5 34182973 - 1 3 1
28 chr1 143910077 - chr1 149459550 - 1 1 0
27 chr22 23632601 + chr9 133729450 + 1 0 0
20 chr12 90313405 - chr21 40684813 - 1 2 0
20 chr22 23632601 + chr9 133655755 + 1 0 1
20 chr9 123636256 - chr9 123578959 + 1 1 4
15 chr16 85589970 + chr6 16762582 + 1 3 2
15 chr3 197348574 - chr3 195392936 + 1 1 0
14 chr18 39584506 + chr18 39560613 - 1 2 0
Note that line 4 and 6 here are BCR/ABL fusions. You would need to filter these junctions further
to see which of them connect known but not homologous genes.

7 Output in transcript coordinates.

With --quantMode TranscriptomeSAM option STAR will output alignments translated into tran-
script coordinates in the Aligned.toTranscriptome.out.bam file (in addition to alignments in ge-
nomic coordinates in Aligned.*.sam/bam files). These transcriptomic alignments can be used with
various transcript quantification software that require reads to be mapped to transcriptome, such as
RSEM or eXpress. For example, RSEM command line would look as follows:
rsem-calculate-expression ... --bam Aligned.toTranscriptome.out.bam
/path/to/RSEM/reference RSEM
Note, that STAR first aligns reads to entire genome, and only then searches for concordance
between alignments and transcripts.This approach offers certain advantages compared to the align-
ment to transcriptome only, by not forcing the alignments to annotated transcripts. Note that
--outFilterMultimapNmax filter only applies to genomic alignments. If an alignment passes this
filter, it is converted to all possible transcriptomic alignments and all of them are output.
By default, the output satisfies RSEM requirements: soft-clipping or indels are not allowed. Use
--quantTranscriptomeBan Singleend to allow insertions, deletions ans soft-clips in the transcrip-
tomic alignments, which can be used by some expression quantification software (e.g. eXpress).

8 Counting number of reads per gene.

With --quantMode GeneCounts option STAR will count number reads per gene while mapping. A
read is counted if it overlaps (1nt or more) one and only one gene. Both ends of the paired-end
read are checked for overlaps. The counts coincide with those produced by htseq-count with default
parameters. This option requires annotations in GTF format (i.e. gene id tag for each exon) specified
in --sjdbGTFfile at the genome generation step or at the mapping step provided in option. STAR
outputs read counts per gene into ReadsPerGene.out.tab file with 4 columns which correspond to
different strandedness options:
column 1: gene ID

column 2: counts for unstranded RNA-seq

18
 column 3: counts for the 1st read strand aligned with RNA (htseq-count option -s yes)

column 4: counts for the 2nd read strand aligned with RNA (htseq-count option -s reverse)

Select the output according to the strandedness of your data. Note, that if you have stranded
data and choose one of the columns 3 or 4, the other column (4 or 3) will give you the
count of antisense reads. With --quantMode TranscriptomeSAM GeneCounts, and get both the
Aligned.toTranscriptome.out.bam and ReadsPerGene.out.tab outputs.

9 2-pass mapping.
For the most sensitive novel junction discovery, it is recommended to run STAR in the 2-pass mode.
It does not significantly increase the number of detected novel junctions, but allows to detect more
splices reads mapping to novel junctions. The basic idea is to run 1st pass of STAR mapping with the
usual parameters, then collect the junctions detected in the first pass, and use them as ”annotated”
junctions for the 2nd pass mapping.

9.1 Multi-sample 2-pass mapping.

For a study with multiple samples, it is recommended to collect 1st pass junctions from all samples.

1. Run 1st mapping pass for all samples with ”usual” parameters. Using annotations is recom-
mended either a the genome generation step, or mapping step.

2. Run 2nd mapping pass for all samples , listing SJ.out.tab files from all samples in
--sjdbFileChrStartEnd /path/to/sj1.tab /path/to/sj2.tab ....

9.2 Per-sample 2-pass mapping.

Annotated junctions will be included in both the 1st and 2nd passes. To run STAR 2-pass mapping for
each sample separately, use --twopassMode Basic option. STAR will perform the 1st pass mapping,
then it will automatically extract junctions, insert them into the genome index, and, finally, re-map
all reads in the 2nd mapping pass. This option can be used with annotations, which can be included
either at the run-time (see #1), or at the genome generation step.
--twopass1readsN defines the number of reads to be mapped in the 1st pass. The default and
most sensitive approach is to set it to -1 (or make it bigger than the number of reads in the sample) -
in which case all reads in the input read file(s) are used in the 1st pass. While it can reduce mapping
time by ∼ 40%, it is not recommended to use a small portion of the reads in the 1st step, since it
will significantly reduce sensitivity for the low expressed novel junctions. The idea to use a portion
of the reads in the 1st pass was inspired by Kim, Langmead and Salzberg in Nature Methods 12,
357–360 (2015).

9.3 2-pass mapping with re-generated genome.

This is the original 2-pass method which involves genome re-generation step in-between 1st and 2nd
passes. Since 2.4.1a, it is recommended to use the on the fly 2-pass options as described above.

19
 1. Run 1st pass STAR for all samples with ”usual” parameters. Genome indices generated with
annotations are recommended.

2. Collect all junctions detected in the 1st pass by merging SJ.out.tab files from all runs. Filter
the junctions by removing likelie false positives, e.g. junctions in the mitochondrion genome,
or non-canonical junctions supported by a few reads. If you are using annotations, only novel
junctions need to be considered here, since annotated junctions will be re-used in the 2nd pass
anyway.

3. Use the filtered list of junctions from the 1st pass with --sjdbFileChrStartEnd option, to-
gether with annotations (via --sjdbGTFfile option) to generate the new genome indices for
the 2nd pass mapping. This needs to be done only once for all samples.

4. Run the 2nd pass mapping for all samples with the new genome index.

10 Merging and mapping of overlapping paired-end reads.

This feature improves mapping accuracy for paired-end libraries with short insert sizes, where many
reads have overlapping mates. Importantly, it allows detection of chimeric junction in the overlap
region.
STAR will search for an overlap between mates larger or equal to --peOverlapNbasesMin bases
with proportion of mismatches in the overlap area not exceeding --peOverlapMMp. If the overlap
is found, STAR will map merge the mates and attempt to map the resulting (single-end) sequence.
If requested, the chimeric detection will be performed on the merged-mate sequence, thus allowing
chimeric detection in the overlap region. If the score of this alignment higher than the original one, or
if a chimeric alignment is found, STAR will report the merged-mate aligment instead of the original
one. In the output, the merged-mate aligment will be converted back to paired-end format.
The developmment of this algorithm was supported by Illumina, Inc. Many thanks to June
Snedecor, Xiao Chen, and Felix Schlesinger for their extensive help in developing this feature.

11 Detection of personal variants overlapping alignments.

Option --varVCFfile /path/to/vcf/file is used to input VCF file with personal variants. Only
single nucleotide variants (SNVs) are supported at the moment. Each variant is expected to have a
genotype with two alleles. To output variants that overlap alignments, vG and vA have to be added
to --outSAMattributes list. SAM attribute vG outputs the genomic coordinate of the variant,
allowing for identification of the variant. SAM attribute vA outputs which allele is detected in the
read: 1 or 2 match one of the genotype alleles, 3 - no match to genotype.

12 WASP filtering of allele specific alignments.

This is re-implementation of the original WASP algorithm by Bryce van de Geijn, Graham McVicker,
Yoav Gilad and Jonathan K Pritchard. Please cite the original WASP paper: Nature Methods 12,
1061–1063 (2015) https://www.nature.com/articles/nmeth.3582. WASP filtering is activated

20
with --waspOutputMode SAMtag, which will add vW tag to the SAM output: vW:i:1 means alignment
passed WASP filtering, and all other values mean it did not pass:
vW:i:2 - multi-mapping read
vW:i:3 - variant base in the read is N (non-ACGT)
vW:i:4 - remapped read did not map
vW:i:5 - remapped read multi-maps
vW:i:6 - remapped read maps to a different locus
vW:i:7 - read overlaps too many variants

13 STARconsensus
STARconsensus allows for mapping RNA-seq reads to consensus genome. It was introduced in STAR
2.7.7a (2020/12/28).
ˆ Provide the VCF file with consensus SNVs and InDels at the genome generation stage with
--genomeTransformVC Variants.vcf --genomeTransformType Haploid. The alternative al-
leles in this VCF will be inserted to the reference genome to create a ”transformed” genome.
Both the genome sequence and transcript/gene annotations are transformed.
ˆ At the mapping stage, the reads will be mapped to the tranformed (consensus) genome. The
quantification in the transformed annotations can be performed with standard --quantMode
TranscriptomeSAM and/or GeneCounts options. If desired, alignments (SAM/BAM) and
spliced junctions (SJ.out.tab) can be transformed back to the original (reference) coordinates
with --genomeTransformOutput SAM and/or SJ. This is useful if downstream processing relies
on reference coordinates.

14 Detection of multimapping chimeras.

Previous STAR chimeric detection algorithm only detected uniquely mapping chimeras, which re-
duced its sensitivity in some cases. The new algorithm can detect and output multimapping chimeras.
Presently, the only output into Chimeric.out.junction is supported. This algorithm is activated with
> 0 value in chimMultimapNmax, which defines the maximum number of chimeric multi-alignments.
The chimMultimapScoreRange (= 1 by default) parameter defines the score range for multi-mapping
chimeras below the best chimeric score, similar to the outFilterMultimapScoreRange parameter for
normal alignments. The chimNonchimScoreDropMin (= 20 by default) defines the threshold trigger-
ing chimeric detection: the drop in the best non-chimeric alignment score with respect to the read
length has to be greater than this value.

15 STARsolo: mapping, demultiplexing and gene quantifi-

cation for single cell RNA-seq
STARsolo is a turnkey solution for analyzing droplet single cell RNA sequencing data (e.g. 10X
Genomics Chromium System) built directly into STAR code. STARsolo inputs the raw FASTQ
reads files, and performs the following operations:

21
 ˆ error correction and demultiplexing of cell barcodes using user-input whitelist
ˆ mapping the reads to the reference genome using the standard STAR spliced read alignment
algorithm
ˆ error correction and collapsing (deduplication) of Unique Molecular Identifiers (UMIa)
ˆ quantification of per-cell gene expression by counting the number of reads per gene
STARsolo output is designed to be a drop-in replacement for 10X CellRanger gene quantification
output. It follows CellRanger logic for cell barcode whitelisting and UMI deduplication, and produces
nearly identical gene counts in the same format. At the same time STARsolo is 10 times faster than
the CellRanger.
STARsolo solo* options can be found in the Section 16.27. For more detailed description, please
see https://github.com/alexdobin/STAR/blob/master/docs/STARsolo.md.

15.1 Feature statistics summaries.

Feature statistics summaries are recorded in the Solo.out/ directory in files <Feature>.stats
where features are those used in the --soloFeatures option, e.g. Gene.stats. The following
metrics are recorded:
nNinBarcode: number of reads with more than 2 Ns in cell barcode (CB)
nUMIhomopolymer: number of reads with homopolymer in CB
nTooMany: not used at the moment
nNoMatch: number of reads with CBs that do not match whitelist even with one mismatch
All of the above reads are discarded from Solo output. Remaining reads are checked for overlap with
features (e.g. genes):
nUnmapped: number of reads unmapped to the genome
nNoFeature: number of reads that map to the genome but do not belong to a feature
nAmbigFeature: number of reads that belong to more than one feature
nAmbigFeatureMultimap: number of reads that belong to more than one feature and are also mul-
timapping to the genome (this is a subset of the nAmbigFeature)
nTooMany: number of reads with ambiguous CB (i.e. CB matches whitelist with one
mismatch but with posterior probability ¡0.95)
nNoExactMatch: number of reads with CB that matches a whitelist barcode with 1 mis-
match, but this whitelist barcode does not get any other reads with exact
matches of CB
All of the reads above are output in feature (e.g. gene) / cell count matrices.
nExactMatch: number of reads with CB that match the whitelist exactly
nMatch: total number of reads that match CB with 0 or 1 mismatches (this is superset
of nExactMatch)
nCellBarcodes: number of distinct CBs detected
nUMIs: number of distinct UMIs detected
These metrics can be grouped into more broad categories:

22
 nNinBarcode+nUMIhomopolymer+nNoMatch+nTooMany+nNoExactMatch = number of reads
with CBs that do not match whitelist.
nUnmapped+nAmbigFeature = number of reads without defined feature (gene)
nMatch = number of reads that are output as solo counts
The three categoties above summed together should be equal to the total number of reads.

16 Description of all options.

For each STAR version, the most up-to-date information about all STAR parameters can be
found in the parametersDefault file in the STAR source directory. The parameters in the
parametersDefault, as well as in the descriptions below, are grouped by function:
Special attention has to be paid to parameters that start with --out*, as they control the
STAR output.
In particular, --outFilter* parameters control the filtering of output alignments which[] you
might want to tweak to fit your needs.
Output of “chimeric” alignments is controlled by --chim* parameters.
Genome generation is controlled by --genome* parameters.
Annotations (splice junction database) are controlled by --sjdb* options at the genome gen-
eration step.
Tweaking --score*, --align*, --seed*, --win* parameters, which requires understanding of
the STAR alignment algorithm, is recommended only for advanced users.
Below, allowed parameter values are typed in magenta, and default values - in blue.

16.1 Parameter Files

--parametersFiles
default: -
string: name of a user-defined parameters file, ”-”: none. Can only be defined
on the command line.

16.2 System
--sysShell
default: -
string: path to the shell binary, preferably bash, e.g. /bin/bash.
-
the default shell is executed, typically /bin/sh. This was reported to
fail on some Ubuntu systems - then you need to specify path to bash.

23
16.3 Run Parameters
--runMode

default: alignReads

string: type of the run.

alignReads
map reads
genomeGenerate
generate genome files
inputAlignmentsFromBAM
input alignments from BAM. Presently only works with –outWigType
and –bamRemoveDuplicates options.
liftOver
lift-over of GTF files (–sjdbGTFfile) between genome assemblies using
chain file(s) from –genomeChainFiles.
soloCellFiltering </path/to/raw/count/dir/>
</path/to/output/prefix>
STARsolo cell filtering (”calling”) without remapping, followed by the
path to raw count directory and output (filtered) prefix

--runThreadN

default: 1

int: number of threads to run STAR

--runDirPerm

default: User RWX

string: permissions for the directories created at the run-time.

User RWX
user-read/write/execute
All RWX
all-read/write/execute (same as chmod 777)

--runRNGseed

default: 777

int: random number generator seed.

24
16.4 Genome Parameters
--genomeDir

default: ./GenomeDir/
string: path to the directory where genome files are stored (for –runMode
alignReads) or will be generated (for –runMode generateGenome)

--genomeLoad

default: NoSharedMemory

string: mode of shared memory usage for the genome files. Only used with
–runMode alignReads.
LoadAndKeep
load genome into shared and keep it in memory after run
LoadAndRemove
load genome into shared but remove it after run
LoadAndExit
load genome into shared memory and exit, keeping the genome in
memory for future runs
Remove
do not map anything, just remove loaded genome from memory
NoSharedMemory
do not use shared memory, each job will have its own private copy of
the genome
--genomeFastaFiles

default: -
string(s): path(s) to the fasta files with the genome sequences, separated by
spaces. These files should be plain text FASTA files, they *cannot* be zipped.

Required for the genome generation (–runMode genomeGenerate). Can also be

used in the mapping (–runMode alignReads) to add extra (new) sequences to
the genome (e.g. spike-ins).

--genomeChainFiles

default: -

string: chain files for genomic liftover. Only used with –runMode liftOver .

--genomeFileSizes

default: 0

25
 uint(s)>0: genome files exact sizes in bytes. Typically, this should not be
defined by the user.
--genomeTransformOutput

default: None

string(s) which output to transform back to original genome

SAM
SAM/BAM alignments
SJ
splice junctions (SJ.out.tab)
None
no transformation of the output
--genomeChrSetMitochondrial

default: chrM M MT
string(s) names of the mitochondrial chromosomes. Presently only used for
STARsolo statisics output/

16.5 Genome Indexing Parameters - only used with –runMode

genomeGenerate
--genomeChrBinNbits

default: 18
int: =log2(chrBin), where chrBin is the size of the bins for genome storage:
each chromosome will occupy an integer number of bins. For a genome with
large number of contigs, it is recommended to scale this parameter as min(18,
log2[max(GenomeLength/NumberOfReferences,ReadLength)]).

--genomeSAindexNbases

default: 14
int: length (bases) of the SA pre-indexing string. Typically between 10 and 15.
Longer strings will use much more memory, but allow faster searches. For small
genomes, the parameter –genomeSAindexNbases must be scaled down to
min(14, log2(GenomeLength)/2 - 1).

--genomeSAsparseD

default: 1
int>0: suffux array sparsity, i.e. distance between indices: use bigger numbers
to decrease needed RAM at the cost of mapping speed reduction

26
--genomeSuffixLengthMax

default: -1
int: maximum length of the suffixes, has to be longer than read length. -1 =
infinite.
--genomeTransformType

default: None

string: type of genome transformation

None
no transformation
Haploid
replace reference alleles with alternative alleles from VCF file (e.g.
consensus allele)
Diploid
create two haplotypes for each chromosome listed in VCF file, for
genotypes 1—2, assumes perfect phasing (e.g. personal genome)

--genomeTransformVCF

default: -

string: path to VCF file for genome transformation

16.6 Splice Junctions Database

--sjdbFileChrStartEnd

default: -
string(s): path to the files with genomic coordinates (chr <tab> start <tab>
end <tab> strand) for the splice junction introns. Multiple files can be
supplied wand will be concatenated.
--sjdbGTFfile

default: -

string: path to the GTF file with annotations

--sjdbGTFchrPrefix

default: -
string: prefix for chromosome names in a GTF file (e.g. ’chr’ for using
ENSMEBL annotations with UCSC genomes)

27
--sjdbGTFfeatureExon
default: exon
string: feature type in GTF file to be used as exons for building transcripts
--sjdbGTFtagExonParentTranscript
default: transcript id
string: GTF attribute name for parent transcript ID (default ”transcript id”
works for GTF files)
--sjdbGTFtagExonParentGene
default: gene id
string: GTF attribute name for parent gene ID (default ”gene id” works for
GTF files)
--sjdbGTFtagExonParentGeneName
default: gene name
string(s): GTF attrbute name for parent gene name
--sjdbGTFtagExonParentGeneType
default: gene type gene biotype
string(s): GTF attrbute name for parent gene type
--sjdbOverhang
default: 100
int>0: length of the donor/acceptor sequence on each side of the junctions,
ideally = (mate length - 1)
--sjdbScore
default: 2
int: extra alignment score for alignments that cross database junctions
--sjdbInsertSave
default: Basic
string: which files to save when sjdb junctions are inserted on the fly at the
mapping step
Basic
only small junction / transcript files
All
all files including big Genome, SA and SAindex - this will create a
complete genome directory

28
16.7 Variation parameters
--varVCFfile

default: -
string: path to the VCF file that contains variation data. The 10th column
should contain the genotype information, e.g. 0/1

16.8 Input Files

--inputBAMfile

default: -
string: path to BAM input file, to be used with –runMode
inputAlignmentsFromBAM

16.9 Read Parameters

--readFilesType

default: Fastx

string: format of input read files

Fastx
FASTA or FASTQ
SAM SE
SAM or BAM single-end reads; for BAM use –readFilesCommand
samtools view
SAM PE
SAM or BAM paired-end reads; for BAM use –readFilesCommand
samtools view
--readFilesSAMattrKeep

default: All
string(s): for –readFilesType SAM SE/PE, which SAM tags to keep in the
output BAM, e.g.: –readFilesSAMtagsKeep RG PL
All
keep all tags
None
do not keep any tags
--readFilesIn

default: Read1 Read2

29
 string(s): paths to files that contain input read1 (and, if needed, read2)
--readFilesManifest
default: -
string: path to the ”manifest” file with the names of read files. The manifest
file should contain 3 tab-separated columns:
paired-end reads: read1 file name tab read2 file name tab read group line.
single-end reads: read1 file name tab - tab read group line.
Spaces, but not tabs are allowed in file names.
If read group line does not start with ID:, it can only contain one ID field, and
ID: will be added to it.
If read group line starts with ID:, it can contain several fields separated by tab,
and all fields will be be copied verbatim into SAM @RG header line.
--readFilesPrefix
default: -
string: prefix for the read files names, i.e. it will be added in front of the strings
in –readFilesIn
--readFilesCommand
default: -
string(s): command line to execute for each of the input file. This command
should generate FASTA or FASTQ text and send it to stdout
For example: zcat - to uncompress .gz files, bzcat - to uncompress .bz2 files, etc.
--readMapNumber
default: -1
int: number of reads to map from the beginning of the file
-1: map all reads
--readMatesLengthsIn
default: NotEqual
string: Equal/NotEqual - lengths of names,sequences,qualities for both mates
are the same / not the same. NotEqual is safe in all situations.
--readNameSeparator
default: /
string(s): character(s) separating the part of the read names that will be
trimmed in output (read name after space is always trimmed)
--readQualityScoreBase
default: 33
int>=0: number to be subtracted from the ASCII code to get Phred quality
score
30
16.10 Read Clipping
--clipAdapterType
default: Hamming

string: adapter clipping type

Hamming
adapter clipping based on Hamming distance, with the number of
mismatches controlled by –clip5pAdapterMMp
CellRanger4
5p and 3p adapter clipping similar to CellRanger4. Utilizes Opal
package by Martin Šošić: https://github.com/Martinsos/opal
None
no adapter clipping, all other clip* parameters are disregarded
--clip3pNbases

default: 0
int(s): number(s) of bases to clip from 3p of each mate. If one value is given, it
will be assumed the same for both mates.
--clip3pAdapterSeq

default: -
string(s): adapter sequences to clip from 3p of each mate. If one value is given,
it will be assumed the same for both mates.
polyA
polyA sequence with the length equal to read length
--clip3pAdapterMMp

default: 0.1
double(s): max proportion of mismatches for 3p adapter clipping for each mate.
If one value is given, it will be assumed the same for both mates.
--clip3pAfterAdapterNbases

default: 0
int(s): number of bases to clip from 3p of each mate after the adapter clipping.
If one value is given, it will be assumed the same for both mates.
--clip5pNbases

default: 0
int(s): number(s) of bases to clip from 5p of each mate. If one value is given, it
will be assumed the same for both mates.
31
16.11 Limits
--limitGenomeGenerateRAM

default: 31000000000

int>0: maximum available RAM (bytes) for genome generation

--limitIObufferSize

default: 30000000 50000000

int>0: max available buffers size (bytes) for input/output, per thread

--limitOutSAMoneReadBytes

default: 100000
int>0: max size of the SAM record (bytes) for one read. Recommended value:
>(2*(LengthMate1+LengthMate2+100)*outFilterMultimapNmax
--limitOutSJoneRead

default: 1000

int>0: max number of junctions for one read (including all multi-mappers)

--limitOutSJcollapsed

default: 1000000

int>0: max number of collapsed junctions

--limitBAMsortRAM

default: 0
int>=0: maximum available RAM (bytes) for sorting BAM. If =0, it will be set
to the genome index size. 0 value can only be used with –genomeLoad
NoSharedMemory option.
--limitSjdbInsertNsj

default: 1000000
int>=0: maximum number of junction to be inserted to the genome on the fly
at the mapping stage, including those from annotations and those detected in
the 1st step of the 2-pass run
--limitNreadsSoft

default: -1

int: soft limit on the number of reads

32
16.12 Output: general
--outFileNamePrefix

default: ./
string: output files name prefix (including full or relative path). Can only be
defined on the command line.
--outTmpDir

default: -
string: path to a directory that will be used as temporary by STAR. All
contents of this directory will be removed!
- the temp directory will default to outFileNamePrefix STARtmp

--outTmpKeep

default: None

string: whether to keep the tempporary files after STAR runs is finished
None
remove all temporary files
All .. keep all files

--outStd
default: Log

string: which output will be directed to stdout (standard out)

Log
log messages
SAM
alignments in SAM format (which normally are output to
Aligned.out.sam file), normal standard output will go into Log.std.out
BAM Unsorted
alignments in BAM format, unsorted. Requires –outSAMtype BAM
Unsorted
BAM SortedByCoordinate
alignments in BAM format, sorted by coordinate. Requires
–outSAMtype BAM SortedByCoordinate
BAM Quant
alignments to transcriptome in BAM format, unsorted. Requires
–quantMode TranscriptomeSAM
--outReadsUnmapped

33
 default: None
string: output of unmapped and partially mapped (i.e. mapped only one mate
of a paired end read) reads in separate file(s).

None
no output
Fastx
output in separate fasta/fastq files, Unmapped.out.mate1/2

--outQSconversionAdd

default: 0
int: add this number to the quality score (e.g. to convert from Illumina to
Sanger, use -31)

--outMultimapperOrder

default: Old 2.4

string: order of multimapping alignments in the output files

Old 2.4
quasi-random order used before 2.5.0
Random
random order of alignments for each multi-mapper. Read mates
(pairs) are always adjacent, all alignment for each read stay together.
This option will become default in the future releases.

16.13 Output: SAM and BAM

--outSAMtype

default: SAM

strings: type of SAM/BAM output

1st word:
BAM
output BAM without sorting
SAM
output SAM without sorting
None
no SAM/BAM output

2nd, 3rd:

34
 Unsorted
standard unsorted
SortedByCoordinate
sorted by coordinate. This option will allocate extra memory for
sorting which can be specified by –limitBAMsortRAM.
--outSAMmode

default: Full

string: mode of SAM output

None
no SAM output
Full
full SAM output
NoQS
full SAM but without quality scores
--outSAMstrandField

default: None

string: Cufflinks-like strand field flag

None
not used
intronMotif
strand derived from the intron motif. This option changes the output
alignments: reads with inconsistent and/or non-canonical introns are
filtered out.
--outSAMattributes

default: Standard
string: a string of desired SAM attributes, in the order desired for the output
SAM. Tags can be listed in any combination/order.

***Presets:
None
no attributes
Standard
NH HI AS nM
All
NH HI AS nM NM MD jM jI MC ch

***Alignment:

35
 NH
number of loci the reads maps to: =1 for unique mappers, >1 for
multimappers. Standard SAM tag.
HI
multiple alignment index, starts with –outSAMattrIHstart (=1 by
default). Standard SAM tag.
AS
local alignment score, +1/-1 for matches/mismateches, score*
penalties for indels and gaps. For PE reads, total score for two mates.
Stadnard SAM tag.
nM
number of mismatches. For PE reads, sum over two mates.
NM
edit distance to the reference (number of mismatched + inserted +
deleted bases) for each mate. Standard SAM tag.
MD
string encoding mismatched and deleted reference bases (see standard
SAM specifications). Standard SAM tag.
jM
intron motifs for all junctions (i.e. N in CIGAR): 0: non-canonical; 1:
GT/AG, 2: CT/AC, 3: GC/AG, 4: CT/GC, 5: AT/AC, 6: GT/AT. If
splice junctions database is used, and a junction is annotated, 20 is
added to its motif value.
jI
start and end of introns for all junctions (1-based).
XS
alignment strand according to –outSAMstrandField.
MC
mate’s CIGAR string. Standard SAM tag.
ch
marks all segment of all chimeric alingments for –chimOutType
WithinBAM output.
cN
number of bases clipped from the read ends: 5’ and 3’

***Variation:
vA
variant allele
vG
genomic coordinate of the variant overlapped by the read.
vW
1 - alignment passes WASP filtering; 2,3,4,5,6,7 - alignment does not
pass WASP filtering. Requires –waspOutputMode SAMtag.

***STARsolo:

36
 CR CY UR UY
sequences and quality scores of cell barcodes and UMIs for the solo*
demultiplexing.
GX GN
gene ID and gene name for unique-gene reads.
gx gn
gene IDs and gene names for unique- and multi-gene reads.
CB UB
error-corrected cell barcodes and UMIs for solo* demultiplexing.
Requires –outSAMtype BAM SortedByCoordinate.
sM
assessment of CB and UMI.
sS
sequence of the entire barcode (CB,UMI,adapter).
sQ
quality of the entire barcode.

***Unsupported/undocumented:

ha
haplotype (1/2) when mapping to the diploid genome. Requires
genome generated with –genomeTransformType Diploid .
rB
alignment block read/genomic coordinates.
vR
read coordinate of the variant.
--outSAMattrIHstart

default: 1
int>=0: start value for the IH attribute. 0 may be required by some
downstream software, such as Cufflinks or StringTie.
--outSAMunmapped

default: None

string(s): output of unmapped reads in the SAM format

1st word:
None
no output
Within
output unmapped reads within the main SAM file (i.e.
Aligned.out.sam)

2nd word:

37
 KeepPairs
record unmapped mate for each alignment, and, in case of unsorted
output, keep it adjacent to its mapped mate. Only affects
multi-mapping reads.
--outSAMorder

default: Paired

string: type of sorting for the SAM output

Paired: one mate after the other for all paired alignments

PairedKeepInputOrder: one mate after the other for all paired alignments, the
order is kept the same as in the input FASTQ files
--outSAMprimaryFlag

default: OneBestScore
string: which alignments are considered primary - all others will be marked
with 0x100 bit in the FLAG
OneBestScore
only one alignment with the best score is primary
AllBestScore
all alignments with the best score are primary
--outSAMreadID

default: Standard

string: read ID record type

Standard
first word (until space) from the FASTx read ID line, removing /1,/2
from the end
Number
read number (index) in the FASTx file

--outSAMmapqUnique

default: 255

int: 0 to 255: the MAPQ value for unique mappers

--outSAMflagOR

default: 0

38
 int: 0 to 65535: sam FLAG will be bitwise OR’d with this value, i.e.
FLAG=FLAG — outSAMflagOR. This is applied after all flags have been set
by STAR, and after outSAMflagAND. Can be used to set specific bits that are
not set otherwise.
--outSAMflagAND

default: 65535
int: 0 to 65535: sam FLAG will be bitwise AND’d with this value, i.e.
FLAG=FLAG & outSAMflagOR. This is applied after all flags have been set
by STAR, but before outSAMflagOR. Can be used to unset specific bits that
are not set otherwise.
--outSAMattrRGline

default: -
string(s): SAM/BAM read group line. The first word contains the read group
identifier and must start with ”ID:”, e.g. –outSAMattrRGline ID:xxx CN:yy
”DS:z z z”.
xxx will be added as RG tag to each output alignment. Any spaces in the tag
values have to be double quoted.

Comma separated RG lines correspons to different (comma separated) input

files in –readFilesIn. Commas have to be surrounded by spaces, e.g.

–outSAMattrRGline ID:xxx , ID:zzz ”DS:z z” , ID:yyy DS:yyyy

--outSAMheaderHD

default: -

strings: @HD (header) line of the SAM header

--outSAMheaderPG

default: -

strings: extra @PG (software) line of the SAM header (in addition to STAR)

--outSAMheaderCommentFile

default: -

string: path to the file with @CO (comment) lines of the SAM header

--outSAMfilter

default: None

39
 string(s): filter the output into main SAM/BAM files
KeepOnlyAddedReferences
only keep the reads for which all alignments are to the extra reference
sequences added with –genomeFastaFiles at the mapping stage.
KeepAllAddedReferences
keep all alignments to the extra reference sequences added with
–genomeFastaFiles at the mapping stage.
--outSAMmultNmax

default: -1
int: max number of multiple alignments for a read that will be output to the
SAM/BAM files. Note that if this value is not equal to -1, the top scoring
alignment will be output first
-1
all alignments (up to –outFilterMultimapNmax) will be output
--outSAMtlen

default: 1

int: calculation method for the TLEN field in the SAM/BAM files
1
leftmost base of the (+)strand mate to rightmost base of the (-)mate.
(+)sign for the (+)strand mate
2
leftmost base of any mate to rightmost base of any mate. (+)sign for
the mate with the leftmost base. This is different from 1 for
overlapping mates with protruding ends
--outBAMcompression

default: 1
int: -1 to 10 BAM compression level, -1=default compression (6?), 0=no
compression, 10=maximum compression
--outBAMsortingThreadN

default: 0
int: >=0: number of threads for BAM sorting. 0 will default to
min(6,–runThreadN).
--outBAMsortingBinsN

default: 50

int: >0: number of genome bins fo coordinate-sorting

40
16.14 BAM processing
--bamRemoveDuplicatesType

default: -
string: mark duplicates in the BAM file, for now only works with (i) sorted
BAM fed with inputBAMfile, and (ii) for paired-end alignments only
-
no duplicate removal/marking
UniqueIdentical
mark all multimappers, and duplicate unique mappers. The
coordinates, FLAG, CIGAR must be identical
UniqueIdenticalNotMulti
mark duplicate unique mappers but not multimappers.
--bamRemoveDuplicatesMate2basesN

default: 0
int>0: number of bases from the 5’ of mate 2 to use in collapsing (e.g. for
RAMPAGE)

16.15 Output Wiggle

--outWigType

default: None
string(s): type of signal output, e.g. ”bedGraph” OR ”bedGraph read1 5p”.
Requires sorted BAM: –outSAMtype BAM SortedByCoordinate .
1st word:
None
no signal output
bedGraph
bedGraph format
wiggle
wiggle format
2nd word:
read1 5p
signal from only 5’ of the 1st read, useful for CAGE/RAMPAGE etc
read2
signal from only 2nd read
--outWigStrand

default: Stranded

41
 string: strandedness of wiggle/bedGraph output
Stranded
separate strands, str1 and str2
Unstranded
collapsed strands
--outWigReferencesPrefix

default: -
string: prefix matching reference names to include in the output wiggle file, e.g.
”chr”, default ”-” - include all references
--outWigNorm

default: RPM

string: type of normalization for the signal

RPM
reads per million of mapped reads
None
no normalization, ”raw” counts

16.16 Output Filtering

--outFilterType

default: Normal

string: type of filtering

Normal
standard filtering using only current alignment
BySJout
keep only those reads that contain junctions that passed filtering into
SJ.out.tab
--outFilterMultimapScoreRange

default: 1

int: the score range below the maximum score for multimapping alignments

--outFilterMultimapNmax

default: 10
int: maximum number of loci the read is allowed to map to. Alignments (all of
them) will be output only if the read maps to no more loci than this value.

42
 Otherwise no alignments will be output, and the read will be counted as
”mapped to too many loci” in the Log.final.out .
--outFilterMismatchNmax

default: 10

int: alignment will be output only if it has no more mismatches than this value.

--outFilterMismatchNoverLmax

default: 0.3
real: alignment will be output only if its ratio of mismatches to *mapped*
length is less than or equal to this value.
--outFilterMismatchNoverReadLmax

default: 1.0
real: alignment will be output only if its ratio of mismatches to *read* length is
less than or equal to this value.
--outFilterScoreMin

default: 0
int: alignment will be output only if its score is higher than or equal to this
value.
--outFilterScoreMinOverLread

default: 0.66
real: same as outFilterScoreMin, but normalized to read length (sum of mates’
lengths for paired-end reads)
--outFilterMatchNmin

default: 0
int: alignment will be output only if the number of matched bases is higher
than or equal to this value.
--outFilterMatchNminOverLread

default: 0.66
real: sam as outFilterMatchNmin, but normalized to the read length (sum of
mates’ lengths for paired-end reads).
--outFilterIntronMotifs

default: None

43
 string: filter alignment using their motifs
None
no filtering
RemoveNoncanonical
filter out alignments that contain non-canonical junctions
RemoveNoncanonicalUnannotated
filter out alignments that contain non-canonical unannotated junctions
when using annotated splice junctions database. The annotated
non-canonical junctions will be kept.
--outFilterIntronStrands

default: RemoveInconsistentStrands

string: filter alignments

RemoveInconsistentStrands
remove alignments that have junctions with inconsistent strands
None
no filtering

16.17 Output splice junctions (SJ.out.tab)

--outSJtype

default: Standard

string: type of splice junction output

Standard
standard SJ.out.tab output
None
no splice junction output

16.18 Output Filtering: Splice Junctions

--outSJfilterReads

default: All

string: which reads to consider for collapsed splice junctions output

All
all reads, unique- and multi-mappers
Unique
uniquely mapping reads only
--outSJfilterOverhangMin

44
 default: 30 12 12 12
4 integers: minimum overhang length for splice junctions on both sides for: (1)
non-canonical motifs, (2) GT/AG and CT/AC motif, (3) GC/AG and CT/GC
motif, (4) AT/AC and GT/AT motif. -1 means no output for that motif
does not apply to annotated junctions

--outSJfilterCountUniqueMin
default: 3 1 1 1
4 integers: minimum uniquely mapping read count per junction for: (1)
non-canonical motifs, (2) GT/AG and CT/AC motif, (3) GC/AG and CT/GC
motif, (4) AT/AC and GT/AT motif. -1 means no output for that motif
Junctions are output if one of outSJfilterCountUniqueMin OR
outSJfilterCountTotalMin conditions are satisfied
does not apply to annotated junctions

--outSJfilterCountTotalMin
default: 3 1 1 1
4 integers: minimum total (multi-mapping+unique) read count per junction
for: (1) non-canonical motifs, (2) GT/AG and CT/AC motif, (3) GC/AG and
CT/GC motif, (4) AT/AC and GT/AT motif. -1 means no output for that
motif
Junctions are output if one of outSJfilterCountUniqueMin OR
outSJfilterCountTotalMin conditions are satisfied
does not apply to annotated junctions

--outSJfilterDistToOtherSJmin
default: 10 0 5 10
4 integers>=0: minimum allowed distance to other junctions’ donor/acceptor

does not apply to annotated junctions

--outSJfilterIntronMaxVsReadN
default: 50000 100000 200000
N integers>=0: maximum gap allowed for junctions supported by 1,2,3,,,N
reads
i.e. by default junctions supported by 1 read can have gaps <=50000b, by 2
reads: <=100000b, by 3 reads: <=200000. by >=4 reads any gap
<=alignIntronMax
does not apply to annotated junctions

45
16.19 Scoring
--scoreGap
default: 0
int: splice junction penalty (independent on intron motif)
--scoreGapNoncan
default: -8
int: non-canonical junction penalty (in addition to scoreGap)
--scoreGapGCAG
default: -4
GC/AG and CT/GC junction penalty (in addition to scoreGap)
--scoreGapATAC
default: -8
AT/AC and GT/AT junction penalty (in addition to scoreGap)
--scoreGenomicLengthLog2scale
default: -0.25
extra score logarithmically scaled with genomic length of the alignment:
scoreGenomicLengthLog2scale*log2(genomicLength)
--scoreDelOpen
default: -2
deletion open penalty
--scoreDelBase
default: -2
deletion extension penalty per base (in addition to scoreDelOpen)
--scoreInsOpen
default: -2
insertion open penalty
--scoreInsBase
default: -2
insertion extension penalty per base (in addition to scoreInsOpen)
--scoreStitchSJshift
default: 1
maximum score reduction while searching for SJ boundaries in the stitching
step

46
16.20 Alignments and Seeding
--seedSearchStartLmax

default: 50
int>0: defines the search start point through the read - the read is split into
pieces no longer than this value
--seedSearchStartLmaxOverLread

default: 1.0
real: seedSearchStartLmax normalized to read length (sum of mates’ lengths
for paired-end reads)
--seedSearchLmax

default: 0
int>=0: defines the maximum length of the seeds, if =0 seed length is not
limited
--seedMultimapNmax

default: 10000
int>0: only pieces that map fewer than this value are utilized in the stitching
procedure
--seedPerReadNmax

default: 1000

int>0: max number of seeds per read

--seedPerWindowNmax

default: 50

int>0: max number of seeds per window

--seedNoneLociPerWindow

default: 10

int>0: max number of one seed loci per window

--seedSplitMin

default: 12

int>0: min length of the seed sequences split by Ns or mate gap

47
--seedMapMin

default: 5

int>0: min length of seeds to be mapped

--alignIntronMin

default: 21
minimum intron size: genomic gap is considered intron if its
length>=alignIntronMin, otherwise it is considered Deletion
--alignIntronMax

default: 0
maximum intron size, if 0, max intron size will be determined by
(2ˆwinBinNbits)*winAnchorDistNbins
--alignMatesGapMax

default: 0
maximum gap between two mates, if 0, max intron gap will be determined by
(2ˆwinBinNbits)*winAnchorDistNbins
--alignSJoverhangMin

default: 5

int>0: minimum overhang (i.e. block size) for spliced alignments

--alignSJstitchMismatchNmax

default: 0 -1 0 0
4*int>=0: maximum number of mismatches for stitching of the splice junctions
(-1: no limit).

(1) non-canonical motifs, (2) GT/AG and CT/AC motif, (3) GC/AG and
CT/GC motif, (4) AT/AC and GT/AT motif.
--alignSJDBoverhangMin

default: 3
int>0: minimum overhang (i.e. block size) for annotated (sjdb) spliced
alignments
--alignSplicedMateMapLmin

default: 0

48
 int>0: minimum mapped length for a read mate that is spliced

--alignSplicedMateMapLminOverLmate

default: 0.66

real>0: alignSplicedMateMapLmin normalized to mate length

--alignWindowsPerReadNmax

default: 10000

int>0: max number of windows per read

--alignTranscriptsPerWindowNmax

default: 100

int>0: max number of transcripts per window

--alignTranscriptsPerReadNmax

default: 10000

int>0: max number of different alignments per read to consider

--alignEndsType

default: Local

string: type of read ends alignment

Local
standard local alignment with soft-clipping allowed
EndToEnd
force end-to-end read alignment, do not soft-clip
Extend5pOfRead1
fully extend only the 5p of the read1, all other ends: local alignment
Extend5pOfReads12
fully extend only the 5p of the both read1 and read2, all other ends:
local alignment
--alignEndsProtrude

default: 0 ConcordantPair
int, string: allow protrusion of alignment ends, i.e. start (end) of the +strand
mate downstream of the start (end) of the -strand mate

1st word: int: maximum number of protrusion bases allowed

49
 2nd word: string:
ConcordantPair
report alignments with non-zero protrusion as concordant pairs
DiscordantPair
report alignments with non-zero protrusion as discordant pairs
--alignSoftClipAtReferenceEnds

default: Yes
string: allow the soft-clipping of the alignments past the end of the
chromosomes
Yes
allow
No
prohibit, useful for compatibility with Cufflinks
--alignInsertionFlush

default: None

string: how to flush ambiguous insertion positions

None
insertions are not flushed
Right
insertions are flushed to the right

16.21 Paired-End reads

--peOverlapNbasesMin

default: 0
int>=0: minimum number of overlap bases to trigger mates merging and
realignment. Specify >0 value to switch on the ”merginf of overlapping mates”
algorithm.
--peOverlapMMp

default: 0.01

real, >=0 & <1: maximum proportion of mismatched bases in the overlap area

50
16.22 Windows, Anchors, Binning
--winAnchorMultimapNmax

default: 50

int>0: max number of loci anchors are allowed to map to

--winBinNbits

default: 16
int>0: =log2(winBin), where winBin is the size of the bin for the
windows/clustering, each window will occupy an integer number of bins.

--winAnchorDistNbins

default: 9
int>0: max number of bins between two anchors that allows aggregation of
anchors into one window
--winFlankNbins

default: 4
int>0: log2(winFlank), where win Flank is the size of the left and right flanking
regions for each window
--winReadCoverageRelativeMin

default: 0.5
real>=0: minimum relative coverage of the read sequence by the seeds in a
window, for STARlong algorithm only.
--winReadCoverageBasesMin

default: 0
int>0: minimum number of bases covered by the seeds in a window , for
STARlong algorithm only.

51
16.23 Chimeric Alignments
--chimOutType

default: Junctions

string(s): type of chimeric output

Junctions
Chimeric.out.junction
SeparateSAMold
output old SAM into separate Chimeric.out.sam file
WithinBAM
output into main aligned BAM files (Aligned.*.bam)
WithinBAM HardClip
(default) hard-clipping in the CIGAR for supplemental chimeric
alignments (default if no 2nd word is present)
WithinBAM SoftClip
soft-clipping in the CIGAR for supplemental chimeric alignments
--chimSegmentMin

default: 0
int>=0: minimum length of chimeric segment length, if ==0, no chimeric
output
--chimScoreMin

default: 0

int>=0: minimum total (summed) score of the chimeric segments

--chimScoreDropMax

default: 20
int>=0: max drop (difference) of chimeric score (the sum of scores of all
chimeric segments) from the read length
--chimScoreSeparation

default: 10
int>=0: minimum difference (separation) between the best chimeric score and
the next one
--chimScoreJunctionNonGTAG

default: -1

int: penalty for a non-GT/AG chimeric junction

52
--chimJunctionOverhangMin

default: 20

int>=0: minimum overhang for a chimeric junction

--chimSegmentReadGapMax

default: 0

int>=0: maximum gap in the read sequence between chimeric segments

--chimFilter

default: banGenomicN

string(s): different filters for chimeric alignments

None
no filtering
banGenomicN
Ns are not allowed in the genome sequence around the chimeric
junction
--chimMainSegmentMultNmax

default: 10
int>=1: maximum number of multi-alignments for the main chimeric segment.
=1 will prohibit multimapping main segments.
--chimMultimapNmax

default: 0

int>=0: maximum number of chimeric multi-alignments

0
use the old scheme for chimeric detection which only considered
unique alignments
--chimMultimapScoreRange

default: 1
int>=0: the score range for multi-mapping chimeras below the best chimeric
score. Only works with –chimMultimapNmax > 1
--chimNonchimScoreDropMin

default: 20

53
 int>=0: to trigger chimeric detection, the drop in the best non-chimeric
alignment score with respect to the read length has to be greater than this value
--chimOutJunctionFormat
default: 0
int: formatting type for the Chimeric.out.junction file
0
no comment lines/headers
1
comment lines at the end of the file: command line and Nreads: total,
unique/multi-mapping

16.24 Quantification of Annotations

--quantMode
default: -
string(s): types of quantification requested
-
none
TranscriptomeSAM
output SAM/BAM alignments to transcriptome into a separate file
GeneCounts
count reads per gene
--quantTranscriptomeBAMcompression
default: 1 1
int: -2 to 10 transcriptome BAM compression level
-2
no BAM output
-1
default compression (6?)
0
no compression
10
maximum compression
--quantTranscriptomeBan
default: IndelSoftclipSingleend
string: prohibit various alignment type
IndelSoftclipSingleend
prohibit indels, soft clipping and single-end alignments - compatible
with RSEM
Singleend
prohibit single-end alignments

54
16.25 2-pass Mapping
--twopassMode

default: None

string: 2-pass mapping mode.

None
1-pass mapping
Basic
basic 2-pass mapping, with all 1st pass junctions inserted into the
genome indices on the fly
--twopass1readsN

default: -1
int: number of reads to process for the 1st step. Use very large number (or
default -1) to map all reads in the first step.

16.26 WASP parameters

--waspOutputMode

default: None
string: WASP allele-specific output type. This is re-implementation of the
original WASP mappability filtering by Bryce van de Geijn, Graham McVicker,
Yoav Gilad & Jonathan K Pritchard. Please cite the original WASP paper:
Nature Methods 12, 1061–1063 (2015),
https://www.nature.com/articles/nmeth.3582 .

SAMtag
add WASP tags to the alignments that pass WASP filtering

16.27 STARsolo (single cell RNA-seq) parameters

--soloType

default: None

string(s): type of single-cell RNA-seq

55
 CB UMI Simple
(a.k.a. Droplet) one UMI and one Cell Barcode of fixed length in
read2, e.g. Drop-seq and 10X Chromium.
CB UMI Complex
multiple Cell Barcodes of varying length, one UMI of fixed length and
one adapter sequence of fixed length are allowed in read2 only (e.g.
inDrop, ddSeq).
CB samTagOut
output Cell Barcode as CR and/or CB SAm tag. No UMI counting.
–readFilesIn cDNA read1 [cDNA read2 if paired-end]
CellBarcode read . Requires –outSAMtype BAM Unsorted [and/or
SortedByCoordinate]
SmartSeq
Smart-seq: each cell in a separate FASTQ (paired- or single-end),
barcodes are corresponding read-groups, no UMI sequences,
alignments deduplicated according to alignment start and end (after
extending soft-clipped bases)

--soloCBwhitelist

default: -
string(s): file(s) with whitelist(s) of cell barcodes. Only –soloType
CB UMI Complex allows more than one whitelist file.
None
no whitelist: all cell barcodes are allowed
--soloCBstart

default: 1

int>0: cell barcode start base

--soloCBlen

default: 16

int>0: cell barcode length

--soloUMIstart

default: 17

int>0: UMI start base

--soloUMIlen

default: 10

56
 int>0: UMI length

--soloBarcodeReadLength

default: 1

int: length of the barcode read

1
equal to sum of soloCBlen+soloUMIlen
0
not defined, do not check
--soloBarcodeMate

default: 0

int: identifies which read mate contains the barcode (CB+UMI) sequence
0
barcode sequence is on separate read, which should always be the last
file in the –readFilesIn listed
1
barcode sequence is a part of mate 1
2
barcode sequence is a part of mate 2
--soloCBposition

default: -

strings(s) position of Cell Barcode(s) on the barcode read.

Presently only works with –soloType CB UMI Complex, and barcodes are
assumed to be on Read2.
Format for each barcode: startAnchor startPosition endAnchor endPosition
start(end)Anchor defines the Anchor Base for the CB: 0: read start; 1: read
end; 2: adapter start; 3: adapter end

start(end)Position is the 0-based position with of the CB start(end) with

respect to the Anchor Base
String for different barcodes are separated by space.

Example: inDrop (Zilionis et al, Nat. Protocols, 2017):

–soloCBposition 0 0 2 -1 3 1 3 8

--soloUMIposition

57
 default: -

string position of the UMI on the barcode read, same as soloCBposition

Example: inDrop (Zilionis et al, Nat. Protocols, 2017):

–soloCBposition 3 9 3 14

--soloAdapterSequence

default: -
string: adapter sequence to anchor barcodes. Only one adapter sequence is
allowed.
--soloAdapterMismatchesNmax

default: 1

int>0: maximum number of mismatches allowed in adapter sequence.

--soloCBmatchWLtype

default: 1MM multi

string: matching the Cell Barcodes to the WhiteList

Exact
only exact matches allowed
1MM
only one match in whitelist with 1 mismatched base allowed. Allowed
CBs have to have at least one read with exact match.
1MM multi
multiple matches in whitelist with 1 mismatched base allowed,
posterior probability calculation is used choose one of the matches.
Allowed CBs have to have at least one read with exact match. This option
matches best with CellRanger 2.2.0
1MM multi pseudocounts
same as 1MM Multi, but pseudocounts of 1 are added to all whitelist
barcodes.
1MM multi Nbase pseudocounts
same as 1MM multi pseudocounts, multimatching to WL is allowed
for CBs with N-bases. This option matches best with CellRanger >=
3.0.0
EditDist 2
allow up to edit distance of 3 fpr each of the barcodes. May include
one deletion + one insertion. Only works with –soloType
CB UMI Complex. Matches to multiple passlist barcdoes are not
allowed. Similar to ParseBio Split-seq pipeline.

58
--soloInputSAMattrBarcodeSeq

default: -
string(s): when inputting reads from a SAM file (–readsFileType SAM SE/PE),
these SAM attributes mark the barcode sequence (in proper order).

For instance, for 10X CellRanger or STARsolo BAMs, use

–soloInputSAMattrBarcodeSeq CR UR .

This parameter is required when running STARsolo with input from SAM.

--soloInputSAMattrBarcodeQual

default: -
string(s): when inputting reads from a SAM file (–readsFileType SAM SE/PE),
these SAM attributes mark the barcode qualities (in proper order).

For instance, for 10X CellRanger or STARsolo BAMs, use

–soloInputSAMattrBarcodeQual CY UY .

If this parameter is ’-’ (default), the quality ’H’ will be assigned to all bases.

--soloStrand

default: Forward

string: strandedness of the solo libraries:

Unstranded
no strand information
Forward
read strand same as the original RNA molecule
Reverse
read strand opposite to the original RNA molecule
--soloFeatures

default: Gene
string(s): genomic features for which the UMI counts per Cell Barcode are
collected

59
 Gene
genes: reads match the gene transcript
SJ
splice junctions: reported in SJ.out.tab
GeneFull
full gene (pre-mRNA): count all reads overlapping genes’ exons and
introns
GeneFull ExonOverIntron
full gene (pre-mRNA): count all reads overlapping genes’ exons and
introns: prioritize 100% overlap with exons
GeneFull Ex50pAS
full gene (pre-RNA): count all reads overlapping genes’ exons and
introns: prioritize >50% overlap with exons. Do not count reads with
100% exonic overlap in the antisense direction.
--soloMultiMappers

default: Unique

string(s): counting method for reads mapping to multiple genes

Unique
count only reads that map to unique genes
Uniform
uniformly distribute multi-genic UMIs to all genes
Rescue
distribute UMIs proportionally to unique+uniform counts ( first
iteration of EM)
PropUnique
distribute UMIs proportionally to unique mappers, if present, and
uniformly if not.
EM
multi-gene UMIs are distributed using Expectation Maximization
algorithm
--soloUMIdedup

default: 1MM All

string(s): type of UMI deduplication (collapsing) algorithm

60
 1MM All
all UMIs with 1 mismatch distance to each other are collapsed (i.e.
counted once).
1MM Directional UMItools
follows the ”directional” method from the UMI-tools by Smith, Heger
and Sudbery (Genome Research 2017).
1MM Directional
same as 1MM Directional UMItools, but with more stringent criteria
for duplicate UMIs
Exact
only exactly matching UMIs are collapsed.
NoDedup
no deduplication of UMIs, count all reads.
1MM CR
CellRanger2-4 algorithm for 1MM UMI collapsing.
--soloUMIfiltering

default: -

string(s) type of UMI filtering (for reads uniquely mapping to genes)

-
basic filtering: remove UMIs with N and homopolymers (similar to
CellRanger 2.2.0).
MultiGeneUMI
basic + remove lower-count UMIs that map to more than one gene.
MultiGeneUMI All
basic + remove all UMIs that map to more than one gene.
MultiGeneUMI CR
basic + remove lower-count UMIs that map to more than one gene,
matching CellRanger > 3.0.0 .

Only works with –soloUMIdedup 1MM CR

--soloOutFileNames

default: Solo.out/ features.tsv barcodes.tsv matrix.mtx

string(s) file names for STARsolo output:

file name prefix gene names barcode sequences cell feature count matrix

--soloCellFilter

default: CellRanger2.2 3000 0.99 10

string(s): cell filtering type and parameters

61
 None
do not output filtered cells
TopCells
only report top cells by UMI count, followed by the exact number of
cells
CellRanger2.2
simple filtering of CellRanger 2.2.

Can be followed by numbers: number of expected cells, robust maximum

percentile for UMI count, maximum to minimum ratio for UMI count

The harcoded values are from CellRanger: nExpectedCells=3000;

maxPercentile=0.99; maxMinRatio=10
EmptyDrops CR
EmptyDrops filtering in CellRanger flavor. Please cite the original
EmptyDrops paper: A.T.L Lun et al, Genome Biology, 20, 63 (2019):
https://genomebiology.biomedcentral.com/articles/10.1186/s13059-
019-1662-y

Can be followed by 10 numeric parameters: nExpectedCells maxPercentile

maxMinRatio indMin indMax umiMin umiMinFracMedian candMaxN FDR
simN
The harcoded values are from CellRanger: 3000 0.99 10 45000 90000 500 0.01
20000 0.01 10000
--soloOutFormatFeaturesGeneField3

default: "Gene Expression"

string(s): field 3 in the Gene features.tsv file. If ”-”, then no 3rd field is output.

--soloCellReadStats

default: None

string: Output reads statistics for each CB

Standard
standard output

62