Eric Coissac
ebbfe35cbc
Extracted `inverser_chaine` into a reusable utility function with docstring and added unit test to ensure correctness.
82 lines
5.0 KiB
Markdown
82 lines
5.0 KiB
Markdown
# MPHF selection — two-phase indexing architecture
|
||
|
||
## Indexing architecture
|
||
|
||
Kmer indexing per partition proceeds in two phases. The separation is necessary because the exact number of unique kmers in a partition is not known until after counting and filtering.
|
||
|
||
### Superkmer vs kmer counts
|
||
|
||
The `SKFileMeta` sidecar written by `SKFileWriter` records `instances` (unique superkmers) and `length_sum` (total nucleotides). A superkmer of length L contains L − k + 1 kmers, so the kmer count per partition can be estimated as `length_sum − instances × (k − 1)`. This is an **overestimate** of unique kmers: two distinct superkmers (different flanking contexts, same minimizer) can share kmers. The exact count of unique kmers is only known after enumerating and deduplicating them.
|
||
|
||
Note: two superkmers sharing a kmer necessarily share the same minimizer and therefore always land in the same partition — no kmer can appear in two different partitions.
|
||
|
||
### Phase 1 — provisional index and spectrum
|
||
|
||
1. Enumerate all kmers from the dereplicated superkmers of the partition.
|
||
2. Build a provisional MPHF over this key set; capacity is pre-allocated from the sidecar estimate (slight overestimate, harmless).
|
||
3. Accumulate counts: for each kmer in each superkmer, `count[MPHF(kmer)] += sk.count()`.
|
||
4. Compute the kmer frequency spectrum (histogram: occurrences → number of kmers).
|
||
5. Apply count filter (e.g. discard singletons). After filtering, the exact number of surviving kmers is known.
|
||
6. Discard the provisional MPHF.
|
||
|
||
### Phase 2 — definitive index
|
||
|
||
Build a new MPHF over the filtered kmer set only, with the exact key count available. This is the persistent per-partition index used for all downstream operations (queries, set operations).
|
||
|
||
---
|
||
|
||
## Candidates
|
||
|
||
**boomphf** (BBHash algorithm, maintained by 10X Genomics):
|
||
|
||
- ~3.7 bits/key; mature crate, used in production bioinformatics (Pufferfish, Piscem)
|
||
- Parallel construction; well-tested with DNA kmer data at scale
|
||
- Drawback: largest space footprint; streaming construction (no exact count needed) was its main differentiator — irrelevant here since exact count is available at phase 2
|
||
|
||
**ptr_hash** (PtrHash algorithm, Groot Koerkamp, SEA 2025):
|
||
|
||
- ~2.4 bits/key; fastest queries (≥2.1× over alternatives, 8–12 ns/key for u64 in tight loops) and fastest construction (≥3.1×)
|
||
- Requires exact key count at construction — available at phase 2
|
||
- Drawback: published February 2025 — very young, no production track record
|
||
|
||
**FMPHGO** (`ph` crate, Beling, ACM JEA 2023):
|
||
|
||
- ~2.1 bits/key — most compact of the three; good query speed; parallelisable construction
|
||
- More established than ptr_hash; actively maintained
|
||
- Works well with overestimated capacity → natural fit for phase 1
|
||
|
||
## MPHF choice per phase
|
||
|
||
**Phase 1** (provisional, discarded after spectrum computation): FMPHGO. Tolerates overestimated capacity, compact, no need to optimise for query speed on a temporary structure.
|
||
|
||
**Phase 2** (persistent, queried repeatedly): open between FMPHGO and ptr_hash. Exact key count is available, so both operate optimally. ptr_hash's query speed advantage (2.1–3.3×) is meaningful for the persistent index but carries the risk of a very young crate. FMPHGO is the conservative default; ptr_hash is worth revisiting once it has broader production use.
|
||
|
||
boomphf is effectively eliminated: its space overhead is the largest and its streaming-construction advantage does not apply here.
|
||
|
||
---
|
||
|
||
## Space at scale
|
||
|
||
For 1 024 partitions × 100 M kmers/partition (phase 2 index, after filtering):
|
||
|
||
| MPHF | bits/key | Total MPHF size |
|
||
|----------|----------|-----------------|
|
||
| boomphf | 3.7 | ~47 GB |
|
||
| ptr_hash | 2.4 | ~31 GB |
|
||
| FMPHGO | 2.1 | ~27 GB |
|
||
|
||
For a human genome at 30× coverage with 1 024 partitions, realistic partition sizes are 3–30 M unique kmers → 1–8 MB per phase-2 MPHF, well within RAM.
|
||
|
||
## On-disk and mmap considerations
|
||
|
||
All three are in-memory structures. Their internal representation is flat bit arrays (no heap pointers), making them serialisable as contiguous byte blobs and mmappable per partition. True zero-copy access would require rkyv integration; the `ph` crate currently uses serde, so loading involves a copy. Given per-partition MPHF sizes of 1–8 MB, the OS page cache handles this transparently — strict zero-copy is a refinement, not a blocker.
|
||
|
||
No established Rust crate provides a natively on-disk MPHF. **SSHash** (Sparse and Skew Hash) is a complete kmer dictionary designed for disk access and is order-preserving (overlapping kmers receive consecutive indices → cache-friendly count access), but it is C++-only and covers more than just the MPHF layer.
|
||
|
||
## Open questions
|
||
|
||
- Confirm actual partition sizes and overestimation factor on representative metagenomic datasets.
|
||
- Revisit ptr_hash for phase 2 once the crate has broader production track record.
|
||
- Assess rkyv integration cost for FMPHGO if true zero-copy mmap becomes necessary for the persistent index.
|
||
- Keep SSHash in mind if the indexing architecture is reconsidered at a higher level.
|