• Allow >2 consensus sequences (#73) (may be buggy)
• Fix a bug regarding cluster read IDs (#85)

• Add SIMD dispatch
• More robust 2-consensus mode (used in longcallD)
• Fix pyabpoa installation error in python 3.13 (#81)

• Fix a score matrix bug in pyabpoa
• Add consensus mode for pyabpoa: most frequent base at each pos

• Fix a backtrack bug in seeding mode (#66)
• Add consensus mode: most frequent base at each pos (-a1)

• Fix a memory allocation bug in v1.5.0 (caused by int overflow when memory is >4G)

• Fix pyabpoa installation on macos
• pip install also works now

Fix an installation bug

• Take quality score in FASTQ format file as weight (-Q)

• Allow to generate multiple consensus sequences from graph alignment (-d)
• Output sequence name by default in MSA format, remove -A/--msa-header

• Use SIMDE to handle SIMD instructions
• Set seeding as disabled by default, -S/--seeding to enable seeding, remove -N/--no-seeding
• Set min. size of POA windows as 500 in seeding mode
• Add -c/--amio-acid to align amino acid input sequences
• Add is_aa parameter in pyabpoa to align amino acid input sequences
• Fix CMakeLists.txt