Skip to content

Latest commit

 

History

History
615 lines (308 loc) · 24.4 KB

CHANGELOG.md

File metadata and controls

615 lines (308 loc) · 24.4 KB

Changelog

All notable changes to the tool_name Docker file.

The format is based on Keep a Changelog.

This project adheres to Semantic Versioning.


[Unreleased]

Fixed

  • Fixed VariantPeptideIdentifier that ORF ID was added before variant IDs.

Changed

  • Changed variant IDs for circRNAs to use backsplicing site instead of listing all exons/introns. #878

[1.4.2] - 2024-06-23

  • Fixed splitFasta that NovelORF peptides coding transcripts not recognized correctly.

[1.4.1] - 2024-05-26

  • Fixed VariantPepidePool that old versions of SeqUtils.molecular_weight don't handle SeqRecord objects. #874

[1.4.0] - 2024-03-27

  • --coding-novel-orf added to callNoncoding and callVariant to call novel ORF peptides from coding transcripts. #659

  • Renamed callNoncoding to callNovelORF and replaced the internal source type Noncoding to NovelORF. #863

[1.3.1] - 2024-03-18

Added:

  • Flag --backsplicing-only added to callVariant to allow only calling noncanonical peptides spanning backsplicing site from circRNA events. #858

[1.3.0] - 2024-03-11

Fixed:

  • Adjacent variants were not merged as MNVs successfully. The function always exited with nothing.

  • Because of the updating to on-disk GTF, the coding transcripts were not generated and saved successfully. filterFasta is the only command affected.

  • Updated splitFasta and summarizeFasta to accept source combinations in --order-source.

  • Fixed parseCIRCexplorer so the exon/intron indices in variant IDs are sorted correctly.

  • Fixed parseVEP to handle insertions in start-inclusion. #840

  • Fixed callVariant of 'no reference out node found'. #842

  • Fixed mergeFasta to remove redundant FASTA header entries. #846

  • Fixed GenomicAnnotation to parse ENSEMBL style UTR correctly. #687

  • Switch to use python's logging package for logging.

  • Print basic summary to callVariant output. #412

Add

  • Added --timeout-seconds to callVariant.

  • Added metadata.json to the index directory for all essential parameters of how a moPepGen index directory is created, including cleavage parameters. #850

  • Added updateIndex #853

  • Output a peptide table for additional information. #833

[1.2.1] - 2023-10-05

Add

  • Added --graph-output-dir to save graph data in json.

Fixed

  • Fixed summarizeFasta that SEC and W2F on fusion peptides are ignored. #789

  • Fixed callVariant that variant_coordinates_to_gene failed when the deletion is end inclusion and it overlaps with the last nucleotide of an exon. #793

  • Fixed splitFasta that CodonReassign and SECT were not able to be grouped. #796

  • Fixed callVariant that in-frame subgraphs not recognized when they are not in variant bubble.

  • Fixed callVariant that peptides are falsely called if the last miscleaved node is missing a downstream cleavage altering variant. #800

  • Fixed TVGNode that get_max_subgraph_id always returns the last subgraph ID. #802

  • Fixed callVariant with altSplice insertion with intronic frameshift variant which is very closed to the end of the subgraph. #803

  • Fixed bruteForce that accepter variants are skipped if the donor transcript has variants with the same coordinate. #810

  • Fixed splitFasta that source and source group order gets overriden by GVF order. #805

  • Fixed summarizeFasta and splitFasta being too slow. #795

  • Fixed splitFasta to use top priority header for additional split

  • Fixed callVariant that some accepter only ORFs maybe included for noncoding fusion transcripts.

  • Fixed callVariant that when filtering variants for a given transcript/fusion/circRNA, coordinates of end inclusion insertions were not interpreted correctly.

  • Fixed bruteForce that selenocysteine not fixed for deletion alt sequence.

  • Fixed callVariant that nodes with fusion being treated as subgraph out or end node incorrectly.

  • Fixed callVariant that upstream cleavage altering variants were affecting checks for whether a node is hybrid in circRNA.

  • Fixed callVariant that two SNV at the same location in circRNA was affecting hybrid node identification.

  • Fixed callVariant. When creating the cleavage graph, when a variant bubble is processed, the downstream node(s) needs to be identified for the next iteration, and only in-frame node should be used. However some nodes can span over two reading frames, so we should check the last reading frame index instead of the first.

  • Fixed callVariant that accepter transcript variants very closed to the breakpoint were skipped.

Added

  • Added support for --group-source for summarizeFasta. #798

[1.2.0] - 2023-08-03

Fixed

  • Fixed that reference source are not recognized. Switched to use upper case values. #758

  • Fixed generateIndex that symlink was not created properly for GTF with --gtf-symlink.

  • Fixed matplotlib warning message #742

  • Fixed parseVEP that insertions not parsed successfully if the location is a single base. #766

  • Fixed callVariant that peptides with upstream cleavage altering mutations were not called. #670

  • Fixed fuzzTest to take multiple CPUs and use temporary directory.

  • Fixed issue in callVariant of circRNA with a SNV being silent in the first loop but not in the second.

  • Fixed issue that circRNA peptide nodes with and without a silent mutation were collapsed. #778

  • Fixed that circRNA peptides that carry indels incompatible with the orf start node should not be called. #780

  • Fixed that hybrid node was not identified if the variant is in the end of an exon. #782

  • Fixed that in circRNA, cleavage gain from upstream node is added to the the wrong ORF. #783

  • Fixed callVariant that fit_into_codon terminated early in fusion transcripts when there donor has a frameshift and accepter has a variant right after the breakpoint. #786

[1.1.0] - 2023-06-28

Fixed

  • Reduced memory usage by mapping the GTF files into memory instead of reading it all at once. #371

[1.0.0] - 2023-06-15

Added

  • Added the support for calling peptides of Selenocysteine terminated. #684

  • Added the support for calling peptides of W > F codon reassignments. #484

  • Added the support for calling peptides with adjacent SNP/INDEL. #691

  • Updated fake.py and bruteForce to handle selenocysteine, W2F and MNV. #689

  • callAltTranslation added to call peptides with alternative translation without any genomic or transcriptomic variations.

  • Enabled summarizeFasta to create bar plot of the summary results.

Fixed

  • Fixed fake that simulated selenocysteine positions could be in introns.

  • Fixed fake that the last exon was picked for A3SS or first exon for A5SS.

  • Fixed fusion with very small intronic insertion. #707

  • In ThreeFrameTVG when aligning variant bubbles and when nodes are merged, variants were not merged correctly.

  • Fixed TVG that indel merged with downstream fusion treated as subgraph out. #708

  • Fixed parseRMATS to handle more complex situations such as exons interjacent between splicing sites and exons spanning over the splicing site. #715, #716, #717, and PR #720

  • Fixed callVariant that failed when there is a SNV very close to the end on a AltSplice insertion. #723

  • Fixed TranscriptAnnotationModel for not recognizing transcripts with mRNA_end_NF correctly. #724

  • Fixed callVariant issue of altSplice insertion carries an intronic indel that goes back to the original reading frame. #726

  • Fixed callVariant to handle deletion that spans over an entire intron. #732

  • Fixed callVariant to skip peptides earlier if they are either too long or too short to significantly improve efficiency. #736

  • Fixed callVariant to handle hypermutated region with a dynamic cutoff. #738

  • Fixed decoyFasta to make it as default to keep cleavage site amino acid residues unmodified. #750

  • Fixed SeqFeature and GTFSeqFeature to remove the definition of strand and use location.strand. #616

  • Refactored util so all functions are accessible. #749

[0.11.5] - 2023-3-5

Fixed

  • Fixed callVariant that the command line argument --max-variants-per-node and --additional-variants-per-misc not passed to it and the default value was always used.

[0.11.4] - 2023-2-23

Fixed

  • Fixed callVariant that variant peptides may get redundant labels with same information (transcript ID and variants). #679

[0.11.3] - 2023-2-9

Fixed

  • filterFasta failed with the new noncoding peptide FASTA header. #675

[0.11.2] - 2023-2-3

Fixed

  • Noncoding peptide headers not parsed successfully by summarizeFasta #672

[0.11.1] - 2023-1-31

Fixed

  • Fixed callVariant that alt splice insertions were treated as stop altering when they are not.

  • Argument --orf-assignment added to callNoncoding to allow choosing the min or max ORF. #667

[0.11.0] - 2023-1-29

Fixed

  • When filtering variants for circRNA, those on fragments that are shorter than 3 nucleotides will not be included. #613

  • When collapsing nodes with the same sequence, global variants (mostly circRNA) are no longer considered when comparing variants, so that nodes with no other variant won't be discarded mistakenly. #619

  • Fixed issue that for hybrid nodes that span over a junction site on a circRNA, the location got lost and caused it to fail to identify whether the node is at least one loop downstream to an ORF start site. #621

  • Fixed issue that circRNA with only 1 nucleotide was causing it to fail to filter variants. #623

  • Fixed issue that peptide nodes on different subgraphs were collapsed after expanding the variant bubble, causing downstream nodes to be unprocessed and resulting in * in the final sequences. #625

  • Fixed issue that the ORF start site position cannot be interpreted when checking whether it is at least one loop away, because it can be off by 1 when converting the location from the gene coordinate to amino acid. #630

  • Fixed issue that the 'CHROM' attribute of GVF metadata not read in correctly. #629

  • Fixed issue that when a frameshift insertion is on a alt splice frameshift substitution (or insertion), the node became disconnected after aligning the variant bubble. #635

  • Fixed issue that when getting the stop altering mutations, location comparison was done incorrectly by 1. #636

  • Node in circRNA missing downstream stop lost mutation called as variant peptide incorrectly. #637

  • Silent mutation not excluded when it is very closed to anther mutation. #638

  • Stop retaining mutation not excluded. #638

  • Fusion with donor breakpoint smaller than 3 causing it fail to run. #633

  • Alt splice insertion recognized as stop altering incorrectly. #640

  • Fix that variants that overlap with last 3 nucleotides of the transcripts causing it to fail. #645

  • Fixed parseREDItools that the ref and alt nucleotides were not set correctly for negative strands. #644

  • Fixed callVariant that hybrid peptide sequences were called from circRNA. #653

  • Fixed callVariant that peptides with variants of in-frame mutation causing deletion/insertion between two cleavage sites were missed. #655

  • Fixed callVariant that when setting the max number of variants per peptide, the number of miscleavages was not used correctly. #657

Changed

  • The transcript trailing peptides (peptides at the end of the transcript sequence) are now excluded for transcripts with the mRNA_end_NF tag and circRNA regardless of it. Otherwise for transcripts (either coding or noncoding) that the mRNA end is confirmed (without the mRNA_end_NF) they are now included in the final FASTA. #649

  • Change from ray back to pathos for parallelization. #643

  • Gene ID of the transcript from which a noncoding peptide is called is added to the FASTA header. #662

[0.10.1] - 2022-11-2

Fixed

  • Transcriptional coordinate to genomic coordinate not converted successfully when it is the last nucleotide of the transcript. #592

  • When creating the peptide cleavage graph, the end nodes of ate variant bubble with alt splice were collapsed with the reference node causing the graph cleavage process terminated too early resulting uncleaved nodes. #597

  • in callVariant when filtering variants associated with the donor transcript, the left breakpoint coordinate not converted successfully if it is the end of the transcript. #598

  • Large deletion caused by alt splice raised the complexity drastically so had to go back to treat alt splice deletion as subgraph again. #600

  • In circRNA, nodes that span over the backsplicing site with variant before the backsplicing site were not recognized correctly. #602

  • Failed to find the downstream exon start when the breakpoint is the first nucleotide before the exon start of a transcript on the negative strand. #603

  • In bruteForce, variants that start at the fusion breakpoint were not excluded. #604

  • Variant peptides on circRNA are missed by callVariant when there are multiple ORF candidates. #606

[0.10.0] - 2022-10-20

Added

  • Added support for fusion, alternative splicing and circRNA in bruteForce.

Fixed

  • Several issues of bruteForce were fixed for fusion, alternative splicing and circRNA to be consistent with callVariant.

  • In ThreeFrameTVG and PeptideVariantGraph, large deletions (for alternative splicing) are no longer treated as subgraphs any more.

  • Fixed the issue that the subgraph_id attributes of TVGNode and PVGNode are lost after nodes are merged. #566

  • When expanding the aligned variant bubble, if the downstream node of the start node has multiple inbond nodes, nucleotides will be taken from the downstream node and added to each upstreams

  • Fixed callVariant that when filtering variants that are compatible with fusion, the breakpoint site were not recognized correctly. #567

  • For `ThreeFrameTVG', when aligning variant bubbles, if the end of the first variant is the start of the next (e.g. alternative splicing events that sharing the same splicing site), the merged bubble will then contain both variants

  • Fixed callVariant that mutations are assigned as stop altering mutation when there is a start codon after it. #568

  • Fixed callVariant that alternative splicing variants were not recognized as stop altering mutation correctly because their reference sequence from GVF is only the first nucleotide. #569

  • Fixed callVariant that nodes being lost after an in-frame subgraph. #573

  • Fixed callVariant that the actual fusion breakpoint was not found correctly when trying to tell whether a novel start site should be considered.

  • Fixed callVariant that variant peptides were called with variants present in one loop but not in another. #576

  • When collapsing the end nodes when creating the peptide cleavage graph, nodes that contains alt splice deletions are now separated from others. #580

  • Fusion not inserted correctly when the breakpoint is intronic. #578

  • When finding start altering variants from a node, wrong right position was used. #583

[0.9.4] - 2022-09-07

Fixed

  • Fixed issue of alternative splicing deletion that starts at the third nucleotide of start codon. Those variants are now skipped. #560

[0.9.3] - 2022-08-26

Fixed

  • Fixed issue that mouse reference genome and proteome were not parsed correctly. #546

  • Fixed issue that stop-lost inframe deletions being missed during node collapsing. #549

  • Fixed issue that variants missed by callVariant when multiple variants are causing the same sequence. #552

  • Fixed issue that cpop_collapsed attribute was not retained after merging so peptides that don't end with cleavage sites were yield. #554

  • Fixed problem caused by N in the reference DNA sequence. #556


[0.9.2] - 2022-07-29

Fixed

  • For circRNA, each reading frame subgraph is now replicated for 3 times in order to catch all variant peptides that read through the junction site. #514

  • For decoyFasta, overlapping decoy sequences are also counted and printed to the stdout when using reverse. #474

Added

  • Enzyme lysN is added. #523

[0.9.1] - 2022-07-27

Added

  • UTIL command line tool validateNoncodingCalling added to validate the output of callNoncoding with bruteForceNoncoding. #524

Fixed

  • parseVEP updated to only parse the first 13 columns of the VEP TSV. #540

[0.9.0] - 2022-07-18

Fixed

  • Fixed issue that peptides caused by start gain mutations are not called by bruteForce. #492

  • Fixed issue that variant peptides caused by start gain mutations from noncoding transcripts are not called in some cases. Stop gain mutations from downstream are also missed in some cases. #495

  • Fixed issue that some variant peptides were not called with in-frame deletions. #515

  • Fixed issue that the shuffled decoy sequences produced by decoyFasta are not reproducible when input FASTA order is changed. #517

  • Fixed issue that stop lost mutations are not recognized correctly. #519

  • Fixed issue that start gain and stop lost mutations before the novel start site were not excluded. #520

  • Fixed issue that frameshifting mutations on the same node of a novel ORF start site and is right after it was not carried over to downstream nodes. #526

  • Fixed issue that stop lost mutations were not recognized for deletions. #527

  • Fixed issue that in-frame deletion stop retaining mutations were not recognized #528

  • Fixed issue that variant coordinates not handled correctly when the cleavage site is contained in the inserted sequence. #52

  • Fixed issue that peptides are not called with overlapping deletions. #531


[0.8.1] - 2022-06-24

Fixed

  • The attribute global_variant is added to the PeptideVariantGraph so for circRNA, if the entire variant peptide is called from an insertion sequence, the variant of circRNA is still going to be added. So the FASTA header will be written out correctly. #490

[0.8.0] - 2022-06-22

Fixed

  • Fixed issue that variants are not filtered correctly on ThreeFrameCVG (some variants are discarded or retained incorrectly). #488

[0.7.2] - 2022-06-20

Changed

  • Rolling back to not filtering candidate variant peptides that overlaps with canonical peptide pool at transcript level in PeptideVariantGraph because accessing the shared memory object is too slow (which could be optimized in the future).

[0.7.1] - 2022-06-18

Changed

  • callVariant optimized so the memory usage and run time is reduced to better handle hyper-mutated regions. #483

  • Switched from pathos to ray because the latter has better support for shared memory.


[0.7.0] - 2022-06-08

Fixed

  • Fixed the issue of long run time for transcripts with hypermutated regions by limiting number of additional variants per miscleavage (--additional-variants-per-misc). Closed #476 via #477

  • Location was missing in variant labels output by parseREDItools. #478


[0.6.1] - 2022-05-31

Fixed

  • Fixed the issue that large insertions, such as retained introns, that start right after the start codon were not converted to end inclusion successfully. #470

[0.6.1] - 2022-05-31

Fixed

  • Deletion was mistreated as insertion when trying to convert to end-inclusion format. This only affects deletions that start on the third nucleotide of CDS. #468

[0.6.0] - 2022-05-28

Fixed

  • callVariant very slow on certain cases when the transcript is large. #464

  • callVariant failed to find start codon for transcript with fusion being the first variant and the donor breakpoint is intronic. #465


[0.5.1] - 2022-05-22

Added

  • An option --denylist is added to filterFasta to accept a FASTA file to exclude peptide sequences from it.

[0.5.0] - 2022-05-13

Changed

  • Fixed callVariant that it failed when the breakpoint of a fusion is right at the end of the start codon because it tried convert to end inclusion. #454

  • Fixed callVariant that no variant peptides are called when the first variant is a fusion. #454


[0.4.2] - 2022-05-11

Changed

  • A warning is raised in parseVEP when tryping to parse a MNV (multi-nucleotide variant) and skip the record instead of raising an error. #447

  • Fixed summarizeFasta that source order of Noncoding was not recognized. #449


[0.4.1] - 2022-04-27

Changed

  • Fixed the problem that in summarizeFasta output the order of variant sources in the same group is not consistent across runs. #428

  • Argument --ignore-missing-source added to summarizeFasta so sources not present in any GVF can be ignored without raising any error. #436

  • In filterFasta, when filter with expression table, changed to filter out peptides smaller than, instead of smaller or equal to, the value of --quant-cutoff.

  • Fixed the issue that in splitFasta, variant sources are not grouped as they are specified by --group-source #439

Added

  • Resources usage including memory, CPU and time is now printed to stdout in the end of all command line programs.

Fixed

  • Fixed issue that --additional-split not recognized properly in splitFasta. #443

[0.4.0] - 2022-03-17

Added

  • Added CLI command summarizeFasta to output a summary table of the variant peptide FASTA file output by callVariant.

Changed

  • Attribute key for transcript ID is fixed from 'TRANSCRIPT' to 'TRANSCRIPT_ID' in circRNA's GVF files output by parseCIRCExplorer to be the same as other GVF files.

  • Genomic position for each record is added to the GVF file output by parseCIRCExplorer.


[0.3.1] - 2022-03-01

Changed

  • Argument parameter --decoy_string_position is changed to --decoy-string-position. #417

[0.3.0] - 2022-02-24

Added

  • Enable filterFasta to filter by number of miscleavages per peptide. #382

  • Added CLI command mergeFasta to merge multiple variant peptide database Fasta files into one. This could be useful when working with multiplexed proteomic experiments such as TMT. #380

  • Added CLI command decoyFasta to generate decoy database by shuffling or reversing each sequence. #386

  • Added parameter --min-coverage-rna to parseREDItools to filter by total RNA reads at a given position. #392

  • Added CLI command encodeFasta to replace the variant peptide headers with UUIDs. The original FASTA headers are stored in a text file together with the UUIDs. This is to make the FASTA header short enough for library search engines. #389

Changed

  • Donor and accepter transcript IDs are now explicitly included in the variant IDs of fusion in both GVFs and variaint peptide FASTA headers. Closed #376 via #377

  • For fusion, callVariant now looks at the entire accepter sequence for potential variant peptides, rather than only the peptides that contains the breakpoint. #377

  • filterFasta updated to support filter by number of miscleavages. #383

  • In parseVEP, chromosome seqname for each record is now read directly from the gene annotation, to avoid the 'chr' prefix issue. #391

  • The --transcript-id-column parameter of parseREDItools is changed to take 1-based index. #392

  • Changed splitDatabase to splitFasta for consistency. #397

  • Updated generateIndex to reduce the size of genomic annotation data and the memory usage when loaded. #395


[0.2.0] - 2022-01-28

Added

  • Multi-threading is enabled for callVariant to run in parallel.

  • CLI command indexGVF added to generate a index file for quickly access variant data from the corresponding GVF file. Noted that this command is not required to run.

Changed

  • To solve the complexity of subgraphs introduced by fusion and especially alternative splicing insertion and substitution, the SubgraphTree class is added to keep the graph-subgraph relationship between nodes.

  • Variant records are now kept on disk rather than reading the entire GVF file(s) into memory, and only the file pointers to variant records are kept in memory. This significantly reduces the memory usage of callVariant.

  • The command line arguments are standardized across all commands, for example '-i/--input-path' for inputs and '-o/--output-path' for outputs.

  • generateIndex is changed to use compressed text format to store genomic annotation, because for some reason that we are not sure, when loading the pickled genomic annotation, the memory usage is almost doubled. #394


[0.1.0-beta.1] - 2021-12-23

Added

  • Initial beta release of moPepGen, with the three-frame graph based algorithm implemented to call noncanoinical peptides.