Extends the entropy annex pipeline to optionally compute family overlap data in a single pass, introducing a memory-mapped binary format and an in-memory accumulator for efficient storage. Adds --min-shared-family to auto-exclude genomes below a threshold during SNP-family distance computation, and --family-overlap to export a shared family matrix as CSV. Centralizes upper-triangle indexing into a shared helper function to eliminate duplication.
Renames the CLI flag from --metric to --distance and introduces eight closed-form SNP distance models with optional Jin-Nei gamma correction. Integrates the ndarray crate for matrix operations and adds relaxed PHYLIP output formatting. Updates architecture and theory documentation to cover the new sparse matrix variants, algorithmic fixes, and distance metric implementations.
Extends the phylogenetic pipeline with exporters for TNT, PhyG, and IQ-TREE that generate executable scripts, cost matrices, and recoded alignments. Adds internal helpers for state indexing, floating-point matrix scaling, and Floyd-Warshall metric closure to satisfy external tool constraints. Adjusts visibility modifiers for sibling iterators and entropy annex structs, and updates design documentation for evolutionary distance metrics.
Introduce the --sankoff flag to generate subsampled calibration bundles and pairwise cost matrices. Expose internal sibling algorithm types as public to support external consumption. Implement YAML, CSV, and FASTA output serialization using serde, gated by a configurable substitution saturation ceiling and requiring the --subsample flag.
Replaces the previous two-pass design with a single-pass reduction that computes per-genome-pair substitution counts, cardinality and base-composition tallies, and transition cost matrices. Introduces internal APIs to derive downstream n×n matrices from reduced data and applies configurable filtering based on SNP ratios and exclusion masks. Adds a fused pipeline entry point aggregating alignment and distance outputs for downstream phylogenetic commands. Adjusts visibility of existing reduction functions to restrict external access while permitting sibling module invocation.
Replace per-family callbacks with a per-layer batch callback that accumulates surviving families into an Arc<Vec<SurvivingFamily>>. Update the alignment reduction routine to iterate over this batched structure, enabling zero-copy sharing for parallel operations and structuring the code for future extensibility.
Introduces `--sibling-stats` and `--sibling-hist` flags to compute and export family-size distributions from pre-built annexes. The new algorithms module implements parallel and sequential scanning routines, while the CLI layer handles CSV export with deduplication and standardized error handling.
Introduces configurable phylogenetic sampling and masking controls via new CLI flags (`--subsample`, `--entropy`, `--exclude-genome`, etc.). Adds a complete SNP pseudo-alignment pipeline featuring entropy-biased Gaussian sampling, post-hoc state masking, and proportional per-layer filtering. Extends the FASTA writer with a `write_plain_record` API for bare-header output without JSON annotations.
Introduces an entropy.pent persisted format and algorithms to compute per-family Shannon entropy across genomic partitions. Refactors IndexCache to use Arc-based layer sharing for efficient iteration and restructures the siblings module with extension traits for annex generation. Adds a --shannon CLI flag to export partition-level entropy metrics to CSV.
Refactors the `SiblingAnnexBuildExt` trait to `SiblingExt` across all imports, implementations, and documentation. Introduces new modules for streaming and batched iteration over sibling layers, along with a two-phase pipeline for resolving minorant families.
Introduces a new `--sibling-annex` CLI flag that triggers the construction of a sibling-count/minorant annex for multi-genome indices. The implementation adds a `FamilyMask` data structure with memory-mapped I/O, enabling lock-free concurrent updates via atomic bitwise operations. Batched processing improves cache locality and parallelism, while helper functions derive family presence and minorant flags dynamically. The feature is exposed through an `IndexCache` extension trait, protected by an exclusive directory lock to prevent index corruption during construction.
Implement an `obipipeline::Throttle` with an RAII guard to acquire and release concurrency slots. Expose new bounded parallel methods on `IndexCache` to process cached layers with a configurable cap. Refactor downstream aggregation logic to use single-pass parallel map-reduce instead of manual collect-map-reduce sequences, enforcing a maximum of 8 concurrent layer scans to bound memory usage.
Refactor phylogenetic algorithms (`neighbor_joining`, `upgma`) to return an explicit `Tree` struct instead of a serialized Newick string. This change makes serialization an explicit step for downstream consumers via the new public `Tree::to_newick()` method, decoupling tree construction from output formatting. The `siblings` module has also been moved to `siblings_old`.
Introduce the `obikphylo` crate to support genome-vs-genome distance matrix computation and phylogenetic tree inference via Neighbor-Joining and UPGMA algorithms. Extend the `obikmer2` CLI with a new `phylo` command that exposes configurable metrics, presence thresholds, and CSV/Newick output options. Refactor `obikindex` to expose trait-based partial aggregation for efficient distance metric finalization without full matrix materialization. Update dependency graphs and remove obsolete storage modules.
Update all matrix type references, imports, and instantiations across
the codebase to use the new PersistentIntMatrix name. The change also
includes standardizing code formatting, such as multi-line statements
and import ordering, without altering any behavioral logic or public
API contracts.
Relocate cardcomp into the siblings directory and update internal import paths to reflect the new hierarchy. Adjust the crate's public API surface by updating module declarations and re-exports. This is a purely structural change with no modifications to logic, data models, or runtime behavior.
This commit partitions the obikindex crate into multiple focused subcrates (obikfilter, obikmerge, obikquery, obikrebuild, obikselect, obikstats, obikdump, and obikidxcache) to reduce coupling and clarify module boundaries. It standardizes error handling across the workspace using OKIError and OKIResult, updates index APIs to support lazy, disk-backed partition access, and migrates NUMA system utilities to a new obisys crate. All modifications are structural, focusing on dependency graph expansion, import path updates, and API surface reorganization without altering core runtime behavior.
The MphfLayer::open signature now accepts only a directory path. Mode configuration is resolved internally via disk-based auto-detection using LayerEvidence, eliminating the need for explicit arguments from callers. All internal query methods and external call sites have been updated accordingly, while TypedLayer::open temporarily retains the unused parameter for future cleanup.
Define a shared Algorithm trait with an associated Output type and a parameterless run(&mut self) method. Refactor PartitionRouter, Dereplicator, Counter, and LayerBuilder to implement the trait, standardizing receivers to &mut self and moving configuration and progress callbacks to dedicated builder setters. Decouple error handling using a generic boxed error type and update workspace dependencies accordingly.
Migrate index state tracking from filesystem sentinel files to an on-disk JSON schema within `index.meta`. The `IndexMeta` struct is now wrapped in an `Arc` with internal locking, exposing only fallible methods for genome and state access. In-memory mutation capabilities have been removed, requiring callers to handle I/O errors explicitly and pass immutable references to downstream components like `PartitionRouter`. Public sentinel constants have been removed from exports.
Moves pipeline state bookkeeping, including sentinel file marking and spectrum persistence, into the algorithms' run and close methods. Introduces a crate-private extension trait to satisfy Rust's orphan rule while implementing construction-only operations on KmerIndex. Updates helper function visibility for cross-crate access and removes the now-empty index_layer module.
This commit renames the `Layer` type to `KmerLayer` throughout the `obikindex` and `obikphylo` crates. All imports, struct fields, function signatures, pattern matches, and instantiation calls have been updated accordingly. The change is a purely structural refactor that tightens type constraints without altering runtime behavior or data models.
Introduce a dedicated LayerBuilder struct to orchestrate parallel layer 0 construction across partitions. The fluent API supports configurable abundance thresholds and intermediate artifact retention. Parallel execution is delegated to a NUMA-aware PartitionRunner scheduler, while progress reporting and stage timing are shifted to the command layer. A new mark_indexed method cleanly separates state tracking from orchestration by generating a completion sentinel.
The layer API has been updated to rename `index` and `index_batch` to `hash` and `hash_batch`, aligning with raw MPHF slot hashing behavior. New public methods have been added to directly expose kmer hashing and iteration, enforcing strict state validation on empty layers. All other modifications are consistent code formatting adjustments without functional changes.
Decouple k-mer counting from the partitioner by introducing a new `Counter` struct. The module exposes a fluent builder API with optional partial file retention, executes partition processing in parallel via Rayon with memory-aware chunk sizing, and integrates thread-safe progress callbacks. Update all callers to use the new counter, simplify test pipelines by removing serialization overhead, and clarify algorithm separation in module documentation.
The command now supports reopening existing indexes instead of failing when the output file exists. Control flow branches between opening an existing index and constructing a new one, moving configuration setup exclusively to the creation path. Directory existence is enforced upfront with proper I/O error propagation. The --force flag retains its original semantics by removing the target directory before proceeding with a fresh build.
Move partition dereplication logic into a dedicated `obikderep` crate that implements a two-phase hash-split/merge strategy with Rayon for parallel processing. Centralize superkmer file path construction in `obilayeredmap` and update dependent crates to use the new pipeline and shared path helpers. Adjust test suites to explicitly invoke the dereplication step.
Invert dependencies by moving partition path primitives to a dedicated module and updating the index crate accordingly. Reshape the partition router to accept a mutable index reference, enabling chainable configuration and resolving lifetime issues with explicit drops. Shift orchestration logic from the index crate to the CLI, replacing monolithic scatter calls with discrete dereplication and counting steps. Introduce a generic progress callback API to decouple rate calculation from UI rendering, and correct file I/O paths to route layer-0 artifacts under the partition index directory.
The `Layer` enum is transformed into a lifecycle-aware state machine with an `Empty` variant representing an unconstructed directory. Read and query operations now explicitly panic when invoked on this state, enforcing explicit progression through `create()` before use. Iterator methods are updated to handle the new variant exhaustively, and module visibility constants are adjusted to support the refactored structure.
Renames `Layer<D>` to `TypedLayer<D>` to establish a distinct typed abstraction. Introduces a new heterogeneous `Layer` enum that unifies count and presence storage with runtime dispatch, delegating operations to the underlying typed variants. Updates cache, index, and phylo consumers to align with the renamed type and new extension traits, preserving full test suite stability while preparing the foundation for multi-partition caching.
Consolidates partition logic, metadata storage, and layer management directly into KmerIndex. Renames obikpartitionner to obikpartition, retaining only PartitionRouter for superkmer routing. Removes intermediate .partition() accessors in favor of direct methods on the index and updates PartitionCache::build to accept &KmerIndex directly. Derives n_partitions from config.n_bits and consolidates k-mer/minimizer sizes into IndexMeta.config. Fixes a regression where PartitionRouter::open incorrectly defaulted to closed.
Update all dependent crates and internal imports to reference the renamed module. Expose configuration parameters and genome metadata through new public accessor methods in the index API, and update documentation comments to align with the new package name.
Update Cargo manifests, dependency paths, and Rust imports across the workspace to reflect the `obikpartitionner` to `obikpartition` rename. Synchronize architecture and implementation documentation with the new module structure. Fix minor syntax issues in test assertions to ensure compilation compatibility. No behavioral or API changes are introduced.
Rename the KmerPartition type to KmerPartitions across obikindex, obikpartitionner, and obikphylo/siblings to reflect an updated data model. Update the Mat enum in siblings/cache.rs to add a SparsePresence variant and simplify opening logic by delegating sparse versus dense detection to PersistentBitMatrix. Apply consistent code formatting, import reordering, and multi-line refactoring throughout the affected modules.
Replaced scattered direct metadata loading with centralized instance methods on `KmerPartition` to guarantee consistent error mapping and legacy recovery. Introduced `StorageKind`, `LayerContent`, and `EvidenceKind` enums alongside lightweight disk-probe methods that inspect file presence without opening heavy data structures. Updated callers across the index, partitioner, and phylo modules to use the new partition API, and added unit tests validating the introspection behavior.
Introduce a design specification outlining performance bottlenecks in layer data access and an agreed-upon implementation direction for caching. Refactor the codebase to centralize index and layer directory path resolution within the partition object, replacing manual string joining and external helper functions with dedicated accessor methods.
Standardizes layer directory path resolution across obikindex, obikpartitionner, and obikphylo by replacing inline string formatting and join calls with the centralized obilayeredmap::layer_dir utility. This refactoring removes redundant path construction logic while preserving existing iteration bounds, control flow, and public APIs.
Bumps core dependencies including ndarray, rand, hashbrown, niffler, ureq, sysinfo, indicatif, lru, and remove_dir_all. Adapts source code to accommodate breaking changes by migrating RNG initialization, adjusting HTTP response handling, and replacing the fs4 crate with standard library file locking. Adds a planning document for query benchmarking and sparse index regression tests.
Restructure the project by moving the standalone compare_sparse utility into an example directory, removing Sankoff parameter configurations and benchmark scripts, updating version control ignores, and expanding the test suite with diagnostic checks and performance benchmarks.
Explicitly sets `default-features = false` for the `obikindex` dependency to ensure no default features are activated during dependency resolution and compilation.
Adapts sibling k-mer tests to the updated canonical API by replacing string-based construction with direct byte conversion. Removes intermediate reverse complement steps and introduces a helper function to compute reverse complement strings directly from canonical k-mers, preserving existing test output while streamlining setup.
Adds a new binary that validates bit-level consistency between dense and sparse K-mer index representations by sampling slots across partitions and reporting discrepancies. Refactors sibling cache matrix initialization into a centralized factory method to simplify control flow and standardize error handling. Introduces diagnostic configurations, benchmark scripts, and an ignored test case to support k-mer resolution analysis.
Introduces --iqtree-min-freq (default 0.001) to treat low-frequency nucleotide states as missing data during IQ-TREE alignment generation when --free-loss is active. This triggers a recoding pass that folds rare states into the missing symbol, followed by non-informative site removal and alphabet recomputation to maintain output consistency. The change also adds Sankoff model configuration files and updates related tests and documentation.
Replaces premature exponentiation-based row normalization with log-sum-exp arithmetic to prevent tiny probabilities from collapsing to exactly zero. This eliminates spurious infinite costs for valid but rare transitions while preserving correct IEEE 754 semantics for genuinely unobserved pairs. Adds explicit guards against NaN in degenerate rows and includes a regression test verifying finite costs for probabilities as low as 1e-200.
Shift monomorphism filtering from the entropy scan layer to a lightweight, annex-only pre-pass. By replacing `Selection::All` with a pre-filtered subset, expensive per-genome resolution is strictly limited to non-monomorphic families. This avoids processing ~98% of minorants that are known to be monomorphic, while preserving positional read speedups for subsequent runs. The change is an internal performance refinement with no public API modifications.
The index packing API now accepts a `sparse` parameter to generate `PersistentSparseBitMatrix` files alongside existing dense matrices. The sibling cache automatically detects this format via an `is_multi.prsb` marker file and routes queries identically to the dense variant. A new `--sparse` CLI flag exposes the option, with tests verifying end-to-end pipeline correctness and storage equivalence.
Implements a compact, row-major sparse bit matrix backed by memory-mapped components, introducing EliasFano, PersistentFixedIntVec, and PersistentRankSelectBitVec primitives for efficient storage and decoding. Adds a BinaryMatrix trait to unify row-level operations across dense and sparse implementations. Corrects edge-case behaviors for zero-width bit storage and cardinality-0 rows. Delivers reduced on-disk size and faster random row access, with column reads remaining dense-only. Test suites and benchmarks are included but currently marked as ignored.
Introduces `--subsample N` and `--shannon` CLI flags to cap retained variable families via proportional reservoir sampling and compute per-family Shannon entropy. Updates the family scanning API to support explicit selection filtering with early-exit optimization, resolving an indexing drift issue in monomorphic layers. Streams entropy metrics for 15-state and 4-nucleotide spaces directly to CSV while maintaining parallel processing across sibling layers.
Eric Coissac