CI
9c30676332
Build pipeline: vsh-ci-template-p9886
Source commit: d0c648fb7e
Source message: Delete src/bgzip directory (#64)
It was moved to toolbox
198 lines
11 KiB
Plaintext
198 lines
11 KiB
Plaintext
```bash
|
|
arriba -h
|
|
```
|
|
|
|
Arriba gene fusion detector
|
|
---------------------------
|
|
Version: 2.4.0
|
|
|
|
Arriba is a fast tool to search for aberrant transcripts such as gene fusions.
|
|
It is based on chimeric alignments found by the STAR RNA-Seq aligner.
|
|
|
|
Usage: arriba [-c Chimeric.out.sam] -x Aligned.out.bam \
|
|
-g annotation.gtf -a assembly.fa [-b blacklists.tsv] [-k known_fusions.tsv] \
|
|
[-t tags.tsv] [-p protein_domains.gff3] [-d structural_variants_from_WGS.tsv] \
|
|
-o fusions.tsv [-O fusions.discarded.tsv] \
|
|
[OPTIONS]
|
|
|
|
-c FILE File in SAM/BAM/CRAM format with chimeric alignments as generated by STAR
|
|
(Chimeric.out.sam). This parameter is only required, if STAR was run with the
|
|
parameter '--chimOutType SeparateSAMold'. When STAR was run with the parameter
|
|
'--chimOutType WithinBAM', it suffices to pass the parameter -x to Arriba and -c
|
|
can be omitted.
|
|
|
|
-x FILE File in SAM/BAM/CRAM format with main alignments as generated by STAR
|
|
(Aligned.out.sam). Arriba extracts candidate reads from this file.
|
|
|
|
-g FILE GTF file with gene annotation. The file may be gzip-compressed.
|
|
|
|
-G GTF_FEATURES Comma-/space-separated list of names of GTF features.
|
|
Default: gene_name=gene_name|gene_id gene_id=gene_id
|
|
transcript_id=transcript_id feature_exon=exon feature_CDS=CDS
|
|
|
|
-a FILE FastA file with genome sequence (assembly). The file may be gzip-compressed. An
|
|
index with the file extension .fai must exist only if CRAM files are processed.
|
|
|
|
-b FILE File containing blacklisted events (recurrent artifacts and transcripts
|
|
observed in healthy tissue).
|
|
|
|
-k FILE File containing known/recurrent fusions. Some cancer entities are often
|
|
characterized by fusions between the same pair of genes. In order to boost
|
|
sensitivity, a list of known fusions can be supplied using this parameter. The list
|
|
must contain two columns with the names of the fused genes, separated by tabs.
|
|
|
|
-o FILE Output file with fusions that have passed all filters.
|
|
|
|
-O FILE Output file with fusions that were discarded due to filtering.
|
|
|
|
-t FILE Tab-separated file containing fusions to annotate with tags in the 'tags' column.
|
|
The first two columns specify the genes; the third column specifies the tag. The
|
|
file may be gzip-compressed.
|
|
|
|
-p FILE File in GFF3 format containing coordinates of the protein domains of genes. The
|
|
protein domains retained in a fusion are listed in the column
|
|
'retained_protein_domains'. The file may be gzip-compressed.
|
|
|
|
-d FILE Tab-separated file with coordinates of structural variants found using
|
|
whole-genome sequencing data. These coordinates serve to increase sensitivity
|
|
towards weakly expressed fusions and to eliminate fusions with low evidence.
|
|
|
|
-D MAX_GENOMIC_BREAKPOINT_DISTANCE When a file with genomic breakpoints obtained via
|
|
whole-genome sequencing is supplied via the -d
|
|
parameter, this parameter determines how far a
|
|
genomic breakpoint may be away from a
|
|
transcriptomic breakpoint to consider it as a
|
|
related event. For events inside genes, the
|
|
distance is added to the end of the gene; for
|
|
intergenic events, the distance threshold is
|
|
applied as is. Default: 100000
|
|
|
|
-s STRANDEDNESS Whether a strand-specific protocol was used for library preparation,
|
|
and if so, the type of strandedness (auto/yes/no/reverse). When
|
|
unstranded data is processed, the strand can sometimes be inferred from
|
|
splice-patterns. But in unclear situations, stranded data helps
|
|
resolve ambiguities. Default: auto
|
|
|
|
-i CONTIGS Comma-/space-separated list of interesting contigs. Fusions between genes
|
|
on other contigs are ignored. Cfontigs can be specified with or without the
|
|
prefix "chr". Asterisks (*) are treated as wild-cards.
|
|
Default: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 X Y AC_* NC_*
|
|
|
|
-v CONTIGS Comma-/space-separated list of viral contigs. Asterisks (*) are treated as
|
|
wild-cards.
|
|
Default: AC_* NC_*
|
|
|
|
-f FILTERS Comma-/space-separated list of filters to disable. By default all filters are
|
|
enabled. Valid values: homologs, low_entropy, isoforms,
|
|
top_expressed_viral_contigs, viral_contigs, uninteresting_contigs,
|
|
non_coding_neighbors, mismatches, duplicates, no_genomic_support,
|
|
genomic_support, intronic, end_to_end, relative_support,
|
|
low_coverage_viral_contigs, merge_adjacent, mismappers, multimappers,
|
|
same_gene, long_gap, internal_tandem_duplication, small_insert_size,
|
|
read_through, inconsistently_clipped, intragenic_exonic,
|
|
marginal_read_through, spliced, hairpin, blacklist, min_support,
|
|
select_best, in_vitro, short_anchor, known_fusions, no_coverage,
|
|
homopolymer, many_spliced
|
|
|
|
-E MAX_E-VALUE Arriba estimates the number of fusions with a given number of supporting
|
|
reads which one would expect to see by random chance. If the expected number
|
|
of fusions (e-value) is higher than this threshold, the fusion is
|
|
discarded by the 'relative_support' filter. Note: Increasing this
|
|
threshold can dramatically increase the number of false positives and may
|
|
increase the runtime of resource-intensive steps. Fractional values are
|
|
possible. Default: 0.300000
|
|
|
|
-S MIN_SUPPORTING_READS The 'min_support' filter discards all fusions with fewer than
|
|
this many supporting reads (split reads and discordant mates
|
|
combined). Default: 2
|
|
|
|
-m MAX_MISMAPPERS When more than this fraction of supporting reads turns out to be
|
|
mismappers, the 'mismappers' filter discards the fusion. Default:
|
|
0.800000
|
|
|
|
-L MAX_HOMOLOG_IDENTITY Genes with more than the given fraction of sequence identity are
|
|
considered homologs and removed by the 'homologs' filter.
|
|
Default: 0.300000
|
|
|
|
-H HOMOPOLYMER_LENGTH The 'homopolymer' filter removes breakpoints adjacent to
|
|
homopolymers of the given length or more. Default: 6
|
|
|
|
-R READ_THROUGH_DISTANCE The 'read_through' filter removes read-through fusions
|
|
where the breakpoints are less than the given distance away
|
|
from each other. Default: 10000
|
|
|
|
-A MIN_ANCHOR_LENGTH Alignment artifacts are often characterized by split reads coming
|
|
from only one gene and no discordant mates. Moreover, the split
|
|
reads only align to a short stretch in one of the genes. The
|
|
'short_anchor' filter removes these fusions. This parameter sets
|
|
the threshold in bp for what the filter considers short. Default: 23
|
|
|
|
-M MANY_SPLICED_EVENTS The 'many_spliced' filter recovers fusions between genes that
|
|
have at least this many spliced breakpoints. Default: 4
|
|
|
|
-K MAX_KMER_CONTENT The 'low_entropy' filter removes reads with repetitive 3-mers. If
|
|
the 3-mers make up more than the given fraction of the sequence, then
|
|
the read is discarded. Default: 0.600000
|
|
|
|
-V MAX_MISMATCH_PVALUE The 'mismatches' filter uses a binomial model to calculate a
|
|
p-value for observing a given number of mismatches in a read. If
|
|
the number of mismatches is too high, the read is discarded.
|
|
Default: 0.010000
|
|
|
|
-F FRAGMENT_LENGTH When paired-end data is given, the fragment length is estimated
|
|
automatically and this parameter has no effect. But when single-end
|
|
data is given, the mean fragment length should be specified to
|
|
effectively filter fusions that arise from hairpin structures.
|
|
Default: 200
|
|
|
|
-U MAX_READS Subsample fusions with more than the given number of supporting reads. This
|
|
improves performance without compromising sensitivity, as long as the
|
|
threshold is high. Counting of supporting reads beyond the threshold is
|
|
inaccurate, obviously. Default: 300
|
|
|
|
-Q QUANTILE Highly expressed genes are prone to produce artifacts during library
|
|
preparation. Genes with an expression above the given quantile are eligible
|
|
for filtering by the 'in_vitro' filter. Default: 0.998000
|
|
|
|
-e EXONIC_FRACTION The breakpoints of false-positive predictions of intragenic events
|
|
are often both in exons. True predictions are more likely to have at
|
|
least one breakpoint in an intron, because introns are larger. If the
|
|
fraction of exonic sequence between two breakpoints is smaller than
|
|
the given fraction, the 'intragenic_exonic' filter discards the
|
|
event. Default: 0.330000
|
|
|
|
-T TOP_N Only report viral integration sites of the top N most highly expressed viral
|
|
contigs. Default: 5
|
|
|
|
-C COVERED_FRACTION Ignore virally associated events if the virus is not fully
|
|
expressed, i.e., less than the given fraction of the viral contig is
|
|
transcribed. Default: 0.050000
|
|
|
|
-l MAX_ITD_LENGTH Maximum length of internal tandem duplications. Note: Increasing
|
|
this value beyond the default can impair performance and lead to many
|
|
false positives. Default: 100
|
|
|
|
-z MIN_ITD_ALLELE_FRACTION Required fraction of supporting reads to report an internal
|
|
tandem duplication. Default: 0.070000
|
|
|
|
-Z MIN_ITD_SUPPORTING_READS Required absolute number of supporting reads to report an
|
|
internal tandem duplication. Default: 10
|
|
|
|
-u Instead of performing duplicate marking itself, Arriba relies on duplicate marking by a
|
|
preceding program using the BAM_FDUP flag. This makes sense when unique molecular
|
|
identifiers (UMI) are used.
|
|
|
|
-X To reduce the runtime and file size, by default, the columns 'fusion_transcript',
|
|
'peptide_sequence', and 'read_identifiers' are left empty in the file containing
|
|
discarded fusion candidates (see parameter -O). When this flag is set, this extra
|
|
information is reported in the discarded fusions file.
|
|
|
|
-I If assembly of the fusion transcript sequence from the supporting reads is incomplete
|
|
(denoted as '...'), fill the gaps using the assembly sequence wherever possible.
|
|
|
|
-h Print help and exit.
|
|
|
|
Code repository: https://github.com/suhrig/arriba
|
|
Get help/report bugs: https://github.com/suhrig/arriba/issues
|
|
User manual: https://arriba.readthedocs.io/
|
|
Please cite: https://doi.org/10.1101/gr.257246.119 |