v0.2.0
What's Changed
- fix: add fgumi-simd-fastq to publish list and verify completeness by @nh13 in #257
- feat(consensus): add EM-Seq methylation-aware consensus calling by @nh13 in #168
- feat(filter): add EM-Seq methylation filters and performance improvements by @nh13 in #169
- feat(zipper): add --restore-unconverted-bases flag for EM-seq by @nh13 in #170
- docs: add EM-Seq guide and clean up redundant docs by @nh13 in #171
- feat(consensus): add --methylation-mode for EM-Seq/TAPs consensus by @nh13 in #172
- feat(filter): add --methylation-mode to filter for conversion fraction by @nh13 in #173
- docs: add TAPs guide and update EM-Seq guide by @nh13 in #174
- feat(simulate): add --methylation-mode to simulate subcommands by @nh13 in #263
- fix(extract): take last
:-separated field as UMI from read name by @nh13 in #264 - feat(zipper): auto-detect BAM input on stdin via BGZF magic by @nh13 in #267
- perf(zipper): speed up --restore-unconverted-bases hot path by @nh13 in #271
- fix(zipper)!: rename sort-key tag from pa to tc to avoid bwa-mem pa:f collision by @nh13 in #270
- fix(group): emit consecutive MI integers 0..N-1 by @nh13 in #273
- perf(io): async prefetch reader with POSIX_FADV_WILLNEED for BAM/FASTQ inputs by @nh13 in #258
- perf(zipper): restore unconverted bases on raw BAM bytes by @nh13 in #274
- feat(compare): add --command preset for per-stage compare defaults by @nh13 in #276
- feat(sort): add --async-reader for prefetch I/O on input BAM by @nh13 in #261
- docs(reorder-buffer): correct insert complexity and add micro-bench by @nh13 in #265
- feat(raw-bam): unified view/editor API + close #272 gaps by @nh13 in #278
- refactor: migrate fgumi to new raw-bam view/editor API by @nh13 in #279
- refactor(raw-bam)!: narrow re-exports + partial free-fn demotion by @nh13 in #280
- test(raw-bam): benchmarks + proptests + edge cases by @nh13 in #281
- perf(raw-bam): follow-up speedups + SIMD nibble2base by @nh13 in #282
- fix(group): cap metric memory with per-thread accumulators (#285) by @nh13 in #287
- refactor(zipper): use RawRecord methods in restore_unconverted_bases_in_raw_record by @nh13 in #286
- refactor(pipeline): PerThreadAccumulator for all per-batch metric collection by @nh13 in #290
- feat(sort): distribute spill files across multiple temp dirs by @nh13 in #284
- feat(raw-bam): additive API surface by @nh13 in #288
- feat(sam): raw-byte record_utils + RawRecordClipper by @nh13 in #289
- refactor(consensus): migrate internals to RawRecord; drop RecordBuf variants by @nh13 in #291
- refactor(lib): migrate commands, storage, and pipeline to raw-byte records by @nh13 in #292
- fix(unified-pipeline): suppress deadlock false positive when all queues empty by @nh13 in #297
- fix(consensus): preserve BA-only orientation for methylation tags by @nh13 in #298
- docs: fix broken doctests in commands modules by @nh13 in #300
- refactor(consensus): rename duplex-filter tier locals to best/worst by @nh13 in #299
- refactor(lib)!: collapse Template/DecodedRecord to raw-only; drop dead RecordBuf helpers by @nh13 in #294
- refactor(test): migrate workspace tests to fgumi_raw_bam SamBuilder by @nh13 in #295
- fix(pipeline): stream rejects to disk in simplex/duplex/codec/correct by @nh13 in #293
- fix(simulate): apply --unmapped-fraction to mapped-reads output by @nh13 in #304
- refactor!: delete vendored bam_codec; migrate simulate to SamBuilder by @nh13 in #296
- perf: migrate RecordBuf hot paths to raw-byte accessors by @nh13 in #305
- fix(metrics): use htsjdk-compatible Murmur3 for downsample selection by @nh13 in #306
- feat(compare): restore --command preset for per-stage compare defaults by @nh13 in #307
- fix(compare): preserve /A /B strand suffix in paired-UMI MI keys by @nh13 in #308
- chore: release v0.2.0 by @nh13 in #259
- fix(ci): exclude non-publishable crates from publish-workflow guard by @nh13 in #310
- fix(raw-bam): gate cigar_op_kind behind noodles feature by @nh13 in #311
Full Changelog: v0.1.3...v0.2.0