CHANGELOG.md

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[v0.4.1]

Adds

• [docs] Fix documentation links
• [dmr] Address bug where assert fires when one sample does not have any records over a region, fixes #265A

Fixes

• [sample-probs] Add arguments to specify the colors of the histogram output, also make them deterministic when not given
• [summary] Make table ordering deterministic.

[v0.4.0]

Adds

• [motif] Add search and evaluate subcommands under motif command hierarchy.
• [stats, localize] Add stats and localize commands, see documentation for details.
• [dmr, multi] Combine samples when they have the same name.
• [extract] Add option to emit bgzf-compressed output.

Fixes

• [validate] Only consider modification codes attributed to the primary sequence base being validated.
• [pileup, extract] Improve iteration over regions when --include-bed is provided.

Changes

• [dmr] Use htslib tbx module for reading tabix index instead of noodles.
• [pileup] Require .fai FASTA index when a reference is provided, load only sections of reference that are necessary.
• [extract] Separate calls and full commands to produce the read calls, and full tables, respectively.
• [validate] Change "any-mod" code from {A, C, G, T} to *.

[v0.3.3]

Fixes

• [sample-probs, summary, pileup] Refactor sampling algorithm so that it will not over-sample reads leading to excessive memory usage.
• [validate] Only use modifications on the canonical base we're validating (in case the modBAM has multiple, e.g. A and C-mods).
• [entropy] Check that reference sequences are the same as in the bam header

[v0.3.2]

Fixes

• [thresholds] OOB panic fix, #244
• [dmr, pair] Allow strand information in regions to be provided. #240
• [dmr, multi] Fix problem when many pairs are provided (#229)

Adds

• [sample-probs] Change output format of probabilities table to make it easier to parse, also change schema. Output HTML documents with nicer tables.
• [ci] Build in Ubuntu-16 due to Centos7 EOL.

[v0.3.1]

Fixes

• [call-mods] Always change model to "explicit", dropped base modification probabilities should not be interpreted as canonical.
• [dmr, segment] Add pseudo-count to avoid -inf in HMM.
• [find-motifs] Fix crash in exhaustive search.

Adds

• [dmr] Allow specification of mod code-to-primary base on the command line with --assign-code.

[v0.3.1rc1]

Fixes

• [find-motifs] Bug where error would be reported when output tables are specified. Fixes #195

Adds

• [extract] Allows streaming read calls table directly to stdout.

[v0.3.0]

Fixes

• [validate] Fix bug where observations were not being balanced correctly.
• [dmr] Revert noodles to 0.50.0, partially fixes #178.

Adds

• [find-motifs] Add command to find motif sequences that are enriched for base modification.
• [entropy] Add command to calculate methylation entropy.
• [dmr] Add --segment option to categorize modified positions into contiguous "same" and "different" groups.
• [pileup] Add --header option to emit a column header.

Deprecations

• [pileup] Change output to always be all-tab delimited, next version will error with --only-tabs flag.

[v0.2.8-rc1]

Fixes

• [dmr] Don't crash when using matched samples, but not all samples have modification counts at a position. Fixes #165.
• [update-tags] Correctly handle implicit tags when there are no probs, i.e. C+h -> C+h.., fixes #160.

[v0.2.7]

Fixes

• [dmr] Header was incorrect with multiple samples
• [pileup] Improve performance when using --include-bed, only process contigs in the BED file.
• [dmr, single-site] When using multiple samples, don't fail a position when one or more samples doesn't have a modification call at that position.
• [extract] Expose queue size to reduce memory usage with long reads.
• [validate] Report number of calls filtered out with thresholds.

[v0.2.6]

Fixes

• [dmr, single-site] Don't require that there are equal numbers of samples for single site DMR with multiple samples. Fixes #140.
• [dmr, pairwise, region] Protect when zero bedmethyl records are found for a region, fixes #146.

Adds

• [validate] Adds on-the-fly filtering of reads by alignment identity and/or alignment length.

[v0.2.5]

Fixes

• [extract] Only emit mapped reads when --region is provided, but still emit unmapped bases in those reads unless --mapped-only is passed.
• [extract] Performance improvement due to better tracking of interval boundaries.
• [repair] Updates the MN tag on repaired records.

Adds

• [dmr, single-site] Refactor dmr pair without regions (i.e. single site analysis) to increase performance.
• [dmr, single-site] Add estimated MAP-based p-value to output.
• [all] Allows BED3 input for all options that use --include-bed. Strand will be assumed to be BOTH (equivalent to '.').
• [extract] Increases the kmer size limit to 50.

[v0.2.5-rc2]

Fixes

• [all] Reads with entirely implicit canonical calls are no longer skipped for "modbase info empty" or similar.

Adds

• adds the --no-implicit-calls flag to modkit update-tags so when changing from the "probability of modification" modes (".", and "") to explicit mode (?) implicit cannonical calls can be skipped. This is important, for example, if the modification calling model did not have the ? mode available.

[v0.2.5-rc1]

Adds

• modkit validate sub-command for rigorous testing of modified base calling models when ground truth labels are known.

Fixes

• [all] Improve performance when using commands on transcriptome reference (or any reference with many sequences less than ~100kb).

[v0.2.4]

Adds

• [extract, adjust-mods, update-tags, call-mods] Parse MN tag in order to use secondary and supplementary alignments.

Fixes

• [all] Improve performance slightly when using short and frequent motifs with --motif option.

[v0.2.3]

Adds

• [dmr, multi] Allow site-level scoring by omitting the --regions argument. Sites will be collected from the input bedMethyl files.
• [dmr] Friendlier handling of missing files and when regions aren't found in the bedMethyl input files.
• [dmr] allow filtering on valid coverage without changing input
• [extract] output "read calls" extract table, a TSV with the base modification calls for every modified position in every read using the same thresholding algorithm as pileup
• [extract] allow filtering of calls by reference motif (like in pileup) as well as BED regions (and exclude regions)

Fixes

• [extract] Improve performance, especially on longer reads.
• [extract] Improve performance with long reads (actually a bug fix)
• [extract] num_soft_clipped_start and num_soft_clipped_end were incorrect on some reverse-mapped reads

[v0.2.3-rc1]

Adds

• [dmr, pair] Allow site-level scoring by omitting the --regions argument. Sites will be collected from the input bedMethyl files.
• [dmr] Friendlier handling of missing files and when regions aren't found in the bedMethyl input files.

Fixes

• [extract] Improve performance, especially on longer reads.

Changes

[v0.2.2]

Adds

• [all] Allow ChEBI codes and more flexible single-character modification codes. E.g. now you can use c for carboxy cytosine or z for something exotic.
• [extract] Adds --kmer-length option to modulate the kmer length on the command line.

Fixes

• [dmr] Automatically select the correct bedMethyl rows when using multiple modified bases (e.g. C-mods and A-mods). Removes assert and replaces it with a warning.

Deprecated

• In the next version pileup output will only have tab delimiters and the --only-tabs flag will be removed.

[v0.2.2-rc1]

Fixes

• [logging] Changes log level from error to debug when a record doesn't have any base modification probabilities.

[v0.2.1]

Adds

• [adjust-mods, summary, pileup, call-mods] Allows asymmetric edge filter (i.e. filter out base modification calls X bases from the start of the reads and Y bases from the ends). Previously, only one parameter was allowed and filtering was symmetric.

Fixes

• [summary, sample-probs] Fixes bug where inferred canonical bases were not being accounted.
• [dmr] Fail when cannot parse any bedMethyl lines from input, log when there are failures to parse lines, fixes #60
• [dmr] Automatically make directory if it doesn't already exist, fixes #61

[v0.2.0]

Adds

• [dmr] Adds modkit dmr command suite for differential methylation exploratory data analysis.
• [pileup-hemi] Adds modkit pileup-hemi for tabulating double-stranded methylation patterns.

Fixes

• [repair] Fixes potential stack overflow when donor and/or acceptor BAMs aren't sorted correctly.

[v0.1.13]

Fixes

• [extract, mod_bam] Potential stack overflow when many reads are skipped.
• [extract] Output inferred canonical calls when the . MM tag is used.
• [adjust-mods] change MM flag to ? and add inferred canonical calls when --edge-filter is applied.

Adds

• [pileup] performs pileup in batches to decrease peak memory usage, configurable with --chunk_size.
• [pileup] Allow users to narrow output to any number of specified motifs (instead of only CpG dinucleotides) with --motif option

Changes

• [extract] adds inferred column

[v0.1.13-rc1]

Fixes

• [extract, mod_bam] Potential stack overflow when many reads are skipped.

[v0.1.12]

Fixes

• [extract] no longer requires an index to use --num-reads, it will just take the first --num-reads reads if no index is found.

Adds

• New command modkit repair allows projecting base modifiction calls from one modBAM onto another where the reads have beem trimmed or clipped.
• adjust-mods, update-tags, sample-probs, summary, call-mods and extract can now be used with standard input/output streams.

[v0.1.11]

Fixes

• [extract] Correctly handles duplex modBAMs when using --include-bed
• Refactors handling of CRAM when sampling records. Fixes #35. Suggests to use BAM.

Adds

• [extract] Output contains ref_mod_strand and modified_primary_base
• [pileup] New option --max-depth allows increasing depth beyond default in hts-lib and sets the default to 8,000 in case it changes in hts-lib.

[v0.1.11-rc1]

Fixes

• [extract] Correctly handles duplex modBAMs when using --include-bed

[v0.1.10]

Fixes

• Don't log message whilst aggregating base modification probabilities (exposed with duplex reads as input).

Changes

• [pileup] Fail fast when bam index doesn't contain any mapped reads.

[v0.1.9]

Changes

• [pileup] When estimating the pass-threshold only use base modification probabilities if the read base is aligned to the reference (don't use soft-clipped and inserts). Use --include-unmapped to use all base modification probabilities.
• [adjust-mods] requires --ignore, --convert, or --edge-filter.

Adds

• [pileup, extract, sample-probs, summary] Allow narrowing of analysis to specific sites with --include-bed.
• [summary, sample-probs] Add --only-mapped flag that will only report on base modification probabilities if they are mapped to the reference.
• [pileup] Allow partitioning counts to separate bedMethyl files based on SAM tags with --partition-tag option.

Fixes

• [adjust-mods, update-tags, call-mods] Panic when failure to parse SAM header, fixes #29.

[v0.1.8]

Changes

• [call-mods] Emit 0 ML value when a mod code is not called, previously the call was omitted (only the called mod was emitted).

Adds

• [adjust-mods, call-mods, pileup, sample-probs, summary, extract] Allow --edge-filter that will remove base modification probabilities at the ends of reads.

Fixes

• [call-mods] --no-filtering flag properly handled.

Deprecated

• [adjust-mods] In order to allow --edge-filter without also performing an ignore or convert, in the next release --ignore will require an explicit argument (h is no longer the default). In this release, when no edge filter, conversion, or ignore is provided, h will be used for ignore, but this behavior will be removed in the next release.

[v0.1.7]

Changes

Adds

- `modkit extract` sub-command that produces a table of per-read base modification probabilities.
- Allow suppression of progress bar with `--suppress-progress`

Fixes

- Don't count skipped reads towards total when sampling.

[v0.1.6]

Changes

• When ignoring a base modification in pileup, calculate pass thresholds on the probabilities after ignoring the specified probability.

Adds

• [pileup] API to specify pass thresholds per base modification (or canonical) with --mod-threshold option.
• New modBAM->modBAM subcommand, call-mods that will apply pass thresholds and clamp base modification probabilities to zero (canonical) or 1.0 for base modifications
• [sample-probs, summary] Allow --ignore to remove a base modification class from the analysis.

[v0.1.5]

Changes

• Table output is default in modkit summary, columns are all_counts/frac and pass_counts/frac instead of counts/frac and filt_counts/frac. See help for details.

Fixes

• The modkit summary command will not use all of the reads in the input modBAM when the --no-filtering flag is used. To disable sampling (and use all of the reads) use --no-sampling. See docs for details.

Adds

• The sample-probs command will now output histograms of the base modification probabilities with the --hist option.
• "Book-style" docs.

[v0.1.4]

Changes

• Implicitly canonical mode . See SAM tags does not require --force-allow-implicit. However no mode at all still does.
• modkit adjust-mods will allow conversion of base modification codes that aren't in the specification (e.g. Z). In order to process these with modkit pileup you must first run modkit adjust-mods in.bam out.bam --convert Z m (for example).
• Per-canonical base thresholds will be automatically inferred, e.g. different thresholds will be chosen for A mods and C mods. This is also true for modkit summary and modkit sample-probs.
• modkit summary will now allow specification of a --region when using an indexed BAM and will sample evenly over the aligned contigs when the region is omitted.

Adds

• modkit summary has --table option that produces a human-readable table.

Deprecated

• modkit summary tab-separated format will not be default in the next release. The table output (currently specified with the --table option will become the default).

Fixes

• Restructuring of the code allows for more parallelism when estimating threshold values, speeding up this process at the start of modkit pileup or modkit summary.

[v0.1.4-rc1]

Changes

• Filter base modification calls that are < the threshold value instead of <= the threshold value.

[v0.1.3]

Adds

• --mask flag in pileup and motif-bed

Fixes

• Default behavior will find CGs in masked regions of the reference.

[v0.1.2]

Adds

• First release.

Fixes