CHANGES.md

v0.1.1

• fixed up variant size-range filtering so that both min_size and max_size can be used at the same time
• will get rid of any inversion with size = 1
• read_support now tallied by Variant instead of Adjacency so that an inversion with 2 breakpoints/adjacencies will have its read_support (of each adjacency) summed up (bwa_mem, no multi-mapping) for filtering

v0.1.2

• Fixed bug in ignoring orientations when creating fusion
• new functions in generating BED track by alignment.py:
  • output correct strand (useful for displaying strand-specific contig alignment)
  • minimum size --min_size can be specified so that small contig alignments can be skipped
• Fixed bug in mis-labelling reciprocal translocations as insertions
• Did not use Pysam AlignedRead.rlen for checking if sequence is chimeric as BWA versions later than 0.7.4 outputs only chimeric portion of sequence other than entire sequence

v0.2.0 (all transcriptome changes)

• reports coverage/depth of all exon-exon junctions assembled in BED format junctions.bed
• reports coverage/depth of reference 5'ref5_jn_depth and 3'ref5_jn_depth junctions in events.tsv for gene fusions and all novel splicing events
• renamed header spanning_reads to support_reads
• accepts supplementary gene annotation GTF --suppl_annot for checking novelty of splicing events
• will not call ITD on homopolymer expansion
• changed event-label of same-gene chimera to actual rearrangement (e.g. a duplication within the same gene will be called dup but not fusion)

v0.3.0

• added support for detecting potential transposable element insertion and repeat expansion and contraction
• changed support read detection workflow
• added test data

v0.4.0

• integrated pavfinder_transcriptome
• changed defaults of subseq_len and probe_len in find_sv_transcriptome.py for detecting events in test data
• changed adjacencies.tsv output from genome to adjacencies.bedpe to agree with transcriptome output
• changed event label repeat_reduction to repeat_contraction
• increased maximum microhomology from 5 to 10 for transcriptome analysis

v0.4.1

• chimera subseq checking improvements/fixes: determine subseq length from fasta file instead of cigar string in BAM file (hard-clip creates problem), multi-map checking requires mapping of entire sequence instead of just checking targets

v0.4.2

• minor changes to tap.py:
  • requires samtools version >= 1.0 because samtools sort command changes
  • creates ruffus history file in output directory
  • explicity index fasta file after merging is done and before analysis scripts are run
  • determine num_proc parameter for running find_sv.py
  • changed name of one of duplicated format_read_pairs() functions
  • fixed bug if suppl_annot is not str

v0.4.3

• changes to handle novel untemplated sequence at ITD breakpoint
  • uses blastn to check for duplication, and does not require end-to-end matching, but requires insertion sequence to be at least 15bp(hard-coded) long; allows bases at edges un-matched (specified by max_novel_length default=10)
  • when parsing partial alignments, allow clipped bases (specified by max_novel_length again)
  • introduce parameter min_dup_size: minimum insertion size to check if it is a duplication (default: 15)
  • changed default --max_novel_len from 20 to 10
• added --only_fusions parameter to only look for fusions
• fixed bug in detecting retained_intron: retained_intron in last match block was not detected.

v0.4.4

• bugfix: if retained_intron happened in first and only alignment block, the variable 'j' will be exposed as un-initialized. Initialize it to 0
• always assume every alignment will have blocks but there are actually cases where Pysam cannot determine alignment blocks. Capture and skip those cases.
• fixed r2c bug in tap when running in whole transcriptome mode
• updated expected output files for tap
• updated refgenes.sorted.gtf.gz and setup.py to copy ensGenes.sorted.gtf.gz

v1.0.0

• added fusion-bloom(Make script) for running RNA-Bloom instead of Trans-ABySS
• modified tap.py to deal with multimatch headers from BBT 2.1.1
• modified tap.py to parse in assembly parameters from cfg file
• bugfix: extract gene names from c2t.bam even when no adjacency/event is identified

v1.1.0

• allows 1 base to differ from canonical splice site (GTAG) in detecting novel donors or acceptors because of possibility of splice-site mutation
• can input genome_bam to detect support level of splice site variant
• search through main annotation to make sure novel event is novel; before only supplementary annotations is used for this purpose, and when no supplementary annotation is provided, an erroneous "novel" event may be reported because wrong transcript is assigned for contig

v1.2.0

• associate novel splicing events with genomic variants given WGS alignments
• bugfix: retained_intron captured by multiple contigs not reported because link attribute is lost after merging adjacencies - fixed
• bugfix: successive retained introns not captured - fixed
• bugfix: remove retained_intron called if encompassing exon (not necessarily having exact boundaries) is found in annotation
• remove redundant novel_acceptor and novel_donor calls of novel_introns

v1.3.0

• added tap2.py, using RNA-Bloom instead of Trans-ABySS for assembly

v1.4.0

• allow fusion flanking_pair support reads to overlap span window (default size 8bp, 4 on either side of breakpoint) so that flanking_pair support of short fragment can also be captured
• changed r2c to run BWA-mem in paired-end mode in tap2.py
• fixed bug in tap2.py extracting RNA-Bloom version when there is stderr message while running rnabloom -v

v1.5.0

• added --min_overhang (default=4) for changing minimum overhang size for gathering spanning reads from r2c in find_sv_transcriptome.py
• minor changes to subseq and probe filtering in fusion filtering: relax NM to 4 and use longest of shorter subseqs for checking
• added Q1 filtering for fusion-bloom and other minor changes (re-order steps, time r2c steps, etc)
• tap.py works with BBTv2.3.2 using biobloommicategorizer now

v1.5.1

• minor fixes/changes in samtools piping in r2c and c2t commands in fusion-bloom
• updates usage for fusion-bloom

v1.6.0

• Fusion-Bloom now expects RNA-Bloom v1.2 to be used, as it reduces redundancy at its final stage. *.transcripts.nr.fa will be expected to be output of the assembly stage.
• Minimap2 replaces BWA MEM as the aligner for r2c
• BugFix for read-through identification - previous assumption of single transcript at both up and downstream junctions removed, as it is not uncommon to have some "minor" transcript in annotation

v1.7.0

• conversion to Python3
• added filter_fasta.py and changed filter_fasta to use it
• updated test data

v1.7.1

• bugfixes and changes related to Fusion-Bloom
  • bugfix: more checkings in is_valid() of Alignment object
  • bugfix: is_valid() call in sv_finder.py
  • bugfix: rare case where adj.size is not int
  • added --nproc to running pavfinder

v1.8.0

• can also examine splice-site mutations from vcf file(s) in addition to genome bam
• fix bug so that skip_exon event will now also be examined for splice-site mutations when genome bam or vcf given
• fix bug in within_utr() reported by issue 15

v1.8.1

• use try..except in processing VCF to prevent crash
• renamed find_genome_splicing_support.py to link_ssv.py
• does not give extra weight to matching outermost exon/block boundaries when calculating score in alignment-transcript mapping
• bugfix: sorting chromosome names
• added blastn as a dependency to INSTALL instructions (for detecting duplication events)

v1.8.2

• updated requirements file for creating conda environment for pavfinder, tap, tap2, and fusion-bloom
• updated expected output files for tap, tap2, and fusion-bloom

v1.8.3

• minor fixes for compatibility with Bioconda

v1.8.4

• separate bash and make files for fusion-bloom for Bioconda recipe

v1.8.5

• added link_ssv.py to setup.py