Refactor batch alignment to support wfmash and improve resource management (#24)

Author: AndreaGuarracino

Cargo.lock

@@ -30,10 +30,10 @@ rayon = "1.10"
 byteorder = "1.5"
 tempfile = "3.8"
 ordered-float = "4.0"
-fastga-rs = { git = "https://github.com/pangenome/fastga-rs", rev = "c0bf620" }
-wfmash-rs = { git = "https://github.com/pangenome/wfmash-rs", rev = "e3fb64ef" }
-ragc-core = { git = "https://github.com/ekg/ragc", rev = "40e5cad" }
-onecode = { git = "https://github.com/pangenome/onecode-rs", rev = "f531f5b" }
+fastga-rs = { git = "https://github.com/pangenome/fastga-rs", rev = "c0bf6202506f51c77e47cf8707e58b7b61e2b621" }
+wfmash-rs = { git = "https://github.com/pangenome/wfmash-rs", rev = "e1207fc50487fb5065f4211d175b3c238fcca167" }
+ragc-core = { git = "https://github.com/ekg/ragc", rev = "40e5cad11cab7d4df07a72d6b16d68c2d60b0742" }
+onecode = { git = "https://github.com/pangenome/onecode-rs", rev = "f531f5b0ff54001a898ec4e0c0c761b2bd0a1f34" }
 chrono = "0.4"
 flate2 = "1.0"
 libc = "0.2"

Cargo.toml

@@ -1,11 +1,11 @@
 # SweepGA
 
-Fast genome alignment with plane sweep filtering and scaffolding. Wraps FastGA aligner and applies plane sweep and other filtering methods to keep the best non-overlapping alignments.
+Fast genome alignment with plane sweep filtering and scaffolding. Wraps wfmash and FastGA aligners and applies plane sweep and other filtering methods to keep the best non-overlapping alignments.
 
 ## What it does
 
 SweepGA can:
-1. **Align FASTA files directly** using integrated FastGA (supports .fa.gz)
+1. **Align FASTA files directly** using integrated wfmash or FastGA (supports .fa.gz)
 2. **Filter existing alignments** from any aligner (wfmash, minimap2, etc.)
 3. **Apply scaffolding/chaining** to merge nearby alignments into syntenic regions
 4. **Output multiple formats**: PAF (text) or .1aln (binary ONE format)
@@ -251,7 +251,7 @@ sweepga alignments.paf -j 50k -m many:many > filtered.paf
 
 **`--check-fastga`** - Check FastGA binary locations and exit (diagnostic)
 
-**`--aligner <ALIGNER>`** - Aligner for FASTA input (default: `fastga`)
+**`--aligner <ALIGNER>`** - Aligner for FASTA input: `wfmash` or `fastga` (default: `wfmash`)
 
 **`-f/--frequency <N>`** - FastGA k-mer frequency threshold
 
@@ -261,7 +261,7 @@ sweepga alignments.paf -j 50k -m many:many > filtered.paf
 
 ### Batch Processing
 
-**`--batch-bytes <SIZE>`** - Maximum index size per batch (e.g., `10G`, `500M`). Partitions genomes to fit disk limits when scratch space is limited.
+**`--batch-bytes <SIZE>`** - Maximum resource usage per batch (e.g., `10G`, `500M`). Partitions genomes into batches. FastGA: limits disk (index ~0.1GB + 12 bytes/bp). Wfmash: limits memory (~0.5GB + 20 bytes/bp).
 
 **`--zstd`** - Compress k-mer index with zstd for ~2x disk savings and faster I/O