Runs FusionInspector, a tool which in silico validates fusion predictions by recovering and re-scoring fusion evidence, from its official Docker image. This app is set up to process the '.fusion_predictions.tsv' file produced by the earlier STAR-Fusion alignment step. It produces a summary file 'inspector.FusionInspector.fusions.abridged.tsv' and a 'final' file with more detailed information.
FusionInspector's GitHub repository and wiki are available at the following URLs: https://github.com/FusionInspector/FusionInspector https://github.com/FusionInspector/FusionInspector/wiki
- The DNA Nexus file ID of a saved FusionInspector Docker image, which should be a compressed '.tar.gz'
- The file IDs for the Read 1 FASTQ file(s), provided as an array. Files must be compressed and end '.fastq.gz' or 'fq.gz'.
- The file IDs for the Read 2 FASTQ file(s), provided as an array. Files must be compressed and end '.fastq.gz' or 'fq.gz'.
- The file IDs for 'known fusions' file(s), provided as an array. Files are expected to end '*.txt'. The known fusion files contain regions which should always have fusion evidence checked by FusionInspector, regardless of whether or not they appear in the STAR-Fusion predictions file. They may include commonly-implicated fusion loci for the condition under investigation.
- Fusions must be written in the format geneA--geneB.
- Known fusions may contain multiple tab-separated columns, but only the first column will be used by the app.
- The file ID of the STAR-Fusion predicted fusions, which ends '.fusion_predictions.tsv'. The fusions will be validated by FusionInspector.
- The file ID of a STAR genome resource, which should be a compressed '.tar.gz' file - from https://data.broadinstitute.org/Trinity/CTAT_RESOURCE_LIB/
- This must have the string "CTAT_lib" in its directory name, in order to be accessed by the script.
- A string value for 'include_trinity', either 'true' if Trinity should be run, or 'false' if it should be skipped.
- The user may pass additional parameters using 'opt_parameters', which should be a space-delimited string. If additional parameters aren't passed, FusionInspector defaults will run. Options available are detailed at the bottom of the page under 'Appendix 1'.
- Downloads all inputs and unzips/untars the STAR genome resource.
- Moves the FASTQ files into 'R1' and 'R2' directories, and known fusions into a 'known_fusion' directory.
- Formats some arguments for FusionInspector: reads, known fusions, and STAR-Fusion predictions.
- Makes minor format corrections to the STAR-Fusion predictions file.
- Loads and runs the FusionInspector Docker image:
- The argument '--include_trinity' will be run in the command if the user selected it.
- Optional parameter '--vis' is run, to produce a HTML report of fusion results.
- Optional parameter '--examine_coding_effect' is run, to produce a tsv file with details of possible coding region impacts of each fusion.
- Optional parameter '--extract_fusion_reads_file' is run, outputting files of fusion-mapping reads.
- Optional parameter '--CPU' is set inside the app, according to the number of threads available in the instance.
- If the user passed any text to the input option 'opt_parameters' these will be appended to the command.
- Production versions of this app will need to point to a controlled Docker image in 'references' on DNAnexus to ensure that the same version is run each time.
- Prefixes output file names with the sample name.
- Uploads the output files to DNA Nexus.
- The following outputs are produced both with and without Trinity being run:
- fi_full: a full set of outputs from FusionInspector, as a tsv file.
- fi_abridged: an abridged version of the FusionInspector output, as a tsv file.
- fi_coding: the abridged FusionInspector output containing additional information about potential coding effect, a tsv file
- fi_html: a HTML of fusion evidence which can be viewed in-browser.
- fi_fusion_r1: read 1s which FusionInspector mapped to the fusions, as a gzipped FASTQ file.
- fi_fusion_r2: read 2s which FusionInspector mapped to the fusions, as a gzipped FASTQ file.
- The following outputs are only produced if 'include_trinity' is set to 'true' at run time:
- fi_trinity_fasta: a FASTA file of de novo assembled transcript sequences.
- fi_trinity_gff: a GFF3 file of reconstructed fusion transcript alignments.
- fi_trinity_bed: a BED file of reconstructed fusion transcript alignments.
Further options available to change in FusionInspector, and obtained by running 'FusionInspector -h' inside an interactive Docker, are as below. For more information about the purpose of each parameter, see the 'help' messages for each parameter at https://github.com/FusionInspector/FusionInspector/blob/master/FusionInspector:
- --min_junction_reads MIN_JUNCTION_READS
- --min_sum_frags MIN_SUM_FRAGS
- --min_novel_junction_support MIN_NOVEL_JUNCTION_SUPPORT
- --min_spanning_frags_only MIN_SPANNING_FRAGS_ONLY
- --require_LDAS REQUIRE_LDAS
- --max_promiscuity MAX_PROMISCUITY
- --min_pct_dom_promiscuity MIN_PCT_DOM_PROMISCUITY
- --min_per_id MIN_PER_ID
- --max_mate_dist MAX_MATE_DIST
- --only_fusion_reads
- --capture_genome_alignments
- --write_intermediate_results
- --cleanup
- --annotate
- --aligner_path ALIGNER_PATH
- --fusion_contigs_only
- --extract_fusion_reads_file EXTRACT_FUSION_READS_FILE
- --no_remove_dups
- --version
- --no_FFPM
- --no_splice_score_boost
- --no_shrink_introns
- --shrink_intron_max_length SHRINK_INTRON_MAX_LENGTH
- --skip_EM
- --incl_microH_expr_brkpt_plots
- --predict_cosmic_like
- --STAR_xtra_params STAR_XTRA_PARAMS
- --no_homology_filter
- --no_annot_filter
- --max_sensitivity
- --extreme_sensitivity
The following options are already set in the app:
- --vis
- --examine_coding_effect
- --extract_fusion_reads_file
- --CPU
The below option is unlikely to be needed with our workflow:
- --samples_file SAMPLES_FILE