Changes

CHANGES LOG
-----------

 - Added .github/dependabot.yml file to auto-update GitHub actions

0.40.0
 - change permissions of STAR working directory to allow deletion by standard pipeline tools
 - add Dockerfile to create container with appropriate tools for basic analyses
 - Updated image labels
 
0.39.0
 - use "samtools index" instead of "cram_index to create cram indexes
 - for bam and cram index creation, use "samtools index -o <index_file>" to explicitly indicate the index file

0.38.0
 - add additional autosome specific coverage threshold to final_output_prep.json

0.37.2
 - change bambi i2b command in stage1 analysis to use new boolean version of --nocall-quality flag

0.37.1
 - amend markdup command chain for duplexseq (nanoseq) analyses
 - updated to use samtools merge instead of bammerge for library merging  

0.37.0
 - add i2b_nocall_qual_switch to stage1 analysis to enable "--nocall-quality=2" to be set for bambi i2b

0.36.0
 - add hs_bwa_executable  parameter to allow different bwa executables for target and nchs alignments (e.g. bwa-mem2 and bwa0_6)
 - add optminpixeldif flag (optical distance for duplicates) to bamstreamingmarkduplicates command (biobambam markdup method)

0.35.0
 - replace scramble with samtools in analysis pipeline templates

0.34.0
 - add calmd to biobambam,samtools and duplexseq markdup methods

0.33.1
 - update CI and tests

0.33.0
 - extend list of preserved auxtags in realignment bamreset (for Duplex-Seq)

0.32.4
 - add supplementary files used in dehumanising CLIMB uploads for ENA (Heron)
 - add maxreadlen flag to bamstreamingduplicates in markdup_biobambam with default of 500

0.32.3
 - force markdup_method:none (and primer_clip_method:no_clip) for untagged phiX and non-consented human split

0.32.2
 - change prefix for SamHaplotag log files to include a trailing '.' (simplifies npg_irods archival)

0.32.1
 - add missing edge to top-level human split template

0.32.0
 - add haplotag processing to stage2 analysis
 - move C2A analysis from bam to cram output stream (to void possible malformed MD tag values)
 - remove intermediate files from duplexseq processing
 - flatten edges arrays (burst sub-arrays) as part of vtfp processing

0.31.0
 - add C2A analysis to stage2 of standard pipelines

0.30.0
 - add optional flags to bwa mem command for Hi-C library stage 2 analyses
 - bam files (and md5, index files) produced using samtools and md5sum instead of bamrecompress

0.29.0
 - renamed markdup_botseq.json markdup_duplexseq.json
 - add markdup_botseq.json

0.28.2
 - move samtools calmd to correct position downstream of coord sort, and make calmd -Q flag (quiet) the default (alignment with markdup_method:none)

0.28.1
 - amend samtools command to "ampliconclip" (previously "amplicon-clip")

0.28.0
 - a new template for cases when marking duplicates is not required, with
   an option to clip primers, which relies on a new feature in samtools

0.27.0
 - check for divide by 0 when calc samtools subsample value
 - drop file check, explictly assume fastq.gz files are gzipped when calc #reads for salmon
 - make the code flow more robust in cases of insufficient reads for salmon
 - unconditionally remove auxtags before adapter clipping when realignment_switch is 1

0.26.0
 - add parameters file for top-up merge
 - functional equivalence: enable selection of markdup method - biobambam (default), samtools or picard

0.25.0
 - add target autosome stats file generation to final_output_prep
 - library merge:  
         cram_write_option (default: use_tears, option use_local)
         updated templates to v2.0 format
 - add options to standard stage2 templates to handle realignment
 
release 0.24.2
 - stage 1 and seq alignment templates: remove fastqcheck file generation
 - stage 1 template: remove generation of a fastq file for index read 

release 0.24.1
 - correct count of minimum reads to run Salmon

release 0.24.0
 - allow format selection for stage1 outputs/stage2 inputs (default: cram)
 - stage1 adapter detection now done in parallel with phix alignment
 - add samtools stats qc check to stage1 analysis
 - stage2 analysis templates amended to allow merging of multiple inputs; application of spatial_filter moved from stage1
 - RNA alignments handle single-end runs
 - new hisat2 alignment method
 - use compressed fastq input for star alignment
 - salmon quantification
     check that there are enough reads in the fastq file inputs
     check if quant file is present before copying it
     connect zip node to quant_genes node to ensure zip waits for salmon
 - post_alignment_realignment.json: new port specification format

release 0.23.0
 - target stats (samtools) addition to final output prep

release 0.22.1
 - add default ("ifnull":"salmon") value for quant_method parameter to fix star/tophat2 tag#0 problem

release 0.22.0
 - new template for minimap2 alignment, with optional post-alignment filtering of secondary and supplementary
     alignments, and minimap2_F_value parameter to allow non-default insert size to be specified (-F flag)
 - subsampling of 10K reads (fastq produced by samtools view -s SEED.FRAC) and production of fastqcheck files
     added to stage1 and stage2 analyses (nchs, single-/paired-end)
 - add lane-level samtools stats to stage1
 - spatial filter application moved to stage2 from stage1
 - amend spatial_filter input/output file names for easier identification
 - empty edge entries recognised as noop - filter in viv.pl, ignore in visualisation, do not apply VTFILE
    prefixes when there are no to and from attributes
 - added -F option to minimap2

release 0.21.0
 - salmon_alignment: copy quant.genes.sf to archive directory
 - star_alignment: assign appropriate suffixes to bamtofastq output files
 - merge: flag change for use with tears 1.2.4

release 0.20.1
 - star_alignment
    correct version number (to 2.0)
    do queryname sort after alignment
 - upgrade realignment template to version 2
 - viv.pl: loosen overly strict checks when using output file nodes with subtype dummy

release 0.20.0
 - phase 1 of new format templates (port spec)
 - replacement of AlignmentFilter and spatial filter with bambi commands
 - now handles XA/Y-split with no target alignment

release 0.19.3
 - fixes for nonconsented human split with no target alignment
     change intermediate OUTFILE node to RAFILE
     remove unneeded bammarkduplicates and parameters only used by sort and duplicate marking

release 0.19.2
 - fix for single-end alignments (bwa aln): specify use_STD[IN|OUT] attributes for EXEC nodes in bwa_aln_se_alignment

release 0.19.1
 - fix split_by_chromosome template (ysplit)

release 0.19.0
 - add new templates for STAR alignment and Salmon
 - Small change in final_output_prep to allow bamsort_cmd to have configurable executable and scramble to have optional embed reference param.
 - add select directive
 - splice/prune operations can be specified in param_vals file
 - new version2 format of port specification in templates
 - correct failure to report errors lower than one VTFILE level

release 0.18.6
 - Y-split fixes
     When processing VTFILE nodes, make sure that parameters are reevaluated locally instead of naively inheriting values
         from the enclosing template (otherwise local changes in values of components of a parameter value may be ignored).
     use bambi for split by chromosome (remove java)
     10-vtfp-vtfile.t - add tests for VTFILE functionality

release 0.18.5
 - Correct logic in vftp.pl for allowing edges without id attributes
 - Reduce verbosity in vtfp.pl
 - Fix handling of splice/prune in vftp.pl where source/destination is STDIN/STOUT
 - Use fopid in place of other substitutions
 - use FindBin in vtfp.pl and viv.pl

release 0.18.4
 - removal of -n 1 flag with seqchksum_merge.pl command in merge_aligned template to maintain current behaviour 
 - logging changes
    correct unneeded warnings about changing use_STDIN/use_STDOUT values for INFILE/OUTFILE/RAFILE node
    recognise JSON::PP:Boolean type (remove warning from log)
    allow edges to have no id values (warnings about undefined values in log files should no longer appear)
 - cope with differing line return behaviour of versions of rev utility (test fix)

release 0.18.3
 - bwa mem flags default changes and additions
    don't use -T by default
    [new flag] use -K 100000000 (reads per thread, large value to make alignment runs replicable)
    [new flag] use -Y to soft clip supplementary alignments instead of hard clipping
 - add ability to splice and prune nodes from the final graph
 - final_output_prep changes to support optional targeted stats files.
 - use threaded bamsormadup in place of bamsort for coord sort
 - add parameters to allow selection of java or bambi (default) implementations of i2b and decode

release 0.18.2
 - propagate any 'AH' fields provided by the aligner when providing full SQ headers (bwa0.7.15)
 - fix bamindexdecoder.json template: add pack option to bamindexdecoder_java_cmd array to handle undefined parameters; change default implementation from "samtools decode" to java BamIndexDecoder

release 0.18.1
- corrected typo (add_item instead of additem) which can cause compilation failure

release 0.18.0
- stage1 analysis : update to bcl_phix_deplex_wtsi_stage1_template; addition of new subgraphs
- added bcftools_genotype_call template for library merge gtcheck qc test
- added facility to allow splicing in of tee(pot) nodes at node outputs at viv.pl runtime
- remove bamcheck and add bam_stats
- added in realignment_wtsi_stage2_humansplit_template which was used to split existing target files from iRODS. Now renamed to realignment_wtsi_humansplit_template.
- changes to merge_final_output_prep to remove extra stats files and allow use of reference for existing stats files
- adding in realignment_wtsi_humansplit_notargetalign_template
- updated bamindexdecoder template (stage1 analysis) to use samtools decode by default (java version still an option)

release 0.17.3
 - update Build.PL and tests to have correct and full set of requirements

release 0.17.2
 - turn off compression of AlignmentFilter output
 - changes to merge_final_output_prep to use tears to stream data into iRODS, generate extra stats,
   use ref in stats file generation and minor tidy up.

release 0.17.1
 - -n 1 flag added to seqchksum_merge.pl command in merge_aligned template to allow different tags in column 1
 - scramble compression
   - up to 7 for final output cram files
   - down to 0 for internal bam streams
 - port naming conventions (IN/OUT pre- and postfixes) adopted in templates and enforced in viv.pl
 - vtfp.pl
    improved error reporting
    refactoring to ensure more consistent/intuitive evaluation of parameter values
    allow specification of "local" parameter substitution (within a specific vtnode)
    added --param_vals --export_param_vals flags
    subst directive attributes ifnull and required added
    remove dead code, review relevance of comments, general tidying
    more tests
 - add tests for viv.pl
 - add (secondary stage) extra split template - remove human and another genome

release 0.17
 - human split with no target align (secondary stage) template introduced
 - initial work for stage one (bcl, adapter, phiX and spatial filter to split processing) using templates
 - bammarkduplicates reintroduced for unaligned file because downstream qc processing relies on presence of markdups_metrics file
 - library cram merging: merge_aligned.json and merge_final_output_prep.json
 - remove potential deadlock by using non-blocking open of STDIN

release 0.16.4
 - add extra branch to teepot command in to stream seqchksum output downstream (instead of using a file as an internal node)

release 0.16.3
 - add tee after seqchksum (bam) output to split output to file and cmp node, to remove deadlock
 - remove -s flag from all cmp commands (improve diagnostics)
 - decrease teepot timeout for bmd_multiway node back to 300

release 0.16.2
 - remove superfluous "-" argumant to bamseqchksum command
 - increase teepot timeout for bmd_multiway node to 50000 (from 500)

release 0.16.1
 - add tempdir parameter and verbose (-v) flag for teepot

release 0.16
 - human split: new alignment_wtsi_stage2_humansplit_template.json, seqchksum_hs.json; addition of subst_params to alignment_common
 - seqchksum comparisons: merge bamseqchksum files for outputs for comparison with initial bam file in seqchksum.json
 - added comparison of cram and bam seqchksum within final_output_prep
 - scramble reference optional in final_output_prep template (reference name passed as a parameter instead of via subgraph_io)
 - realignment templates fixes/amendents: default value for common subst_params file; default to cram input

release 0.15
 - fix construction of alternate hash command to construct sha512primesums512 seqchksum file

release 0.14
 - correct prefix value given to calibration_pu -p flag (used in output file naming - it contained an unwanted ".bam")

release 0.13
 - create cram index files

release 0.12
  -viv.pl
    read/write to/from stdin/stdout
    exec failure of a node's command is now fatal (bug fix)
  
  -vtfp.pl
    updated to use new subst_params format in templates
    multiple -keys/-values pairs on the command-line now produce an array on substitution into the template
    when substition of nested parameters is done, array elements which are themselves arrays are flattened
      (its elements are spliced into the position of the original array); net result is that top-level
      parameter substitutions result in either strings or arrays of strings
  
  -templates
    new subst_params format for production templates
    changes to final_output_prep template to add flexibility to this phase of analysis (e.g. y chrom. split)
    output of seqchksum files for hash type sha512primesums512 added

release 0.09
  -install action should not remove lib directory at target