# AFAP1L2: human–horse sequence comparison

The downloaded human Q8N4X5 sequence (818 residues) and selected horse A0A9L0RQI4 sequence (850 residues) share 88.6% identity among 818 paired residues. Paired coverage is 100.0% of human and 96.2% of horse.

Full alignments, sequence hashes and calculation details: [AFAP1L2.json](../../../../projects/PROTNLM_EVALUATION/mammal-benchmark/paired-sequences/AFAP1L2.json). Reproduce with [compare.py](../../../../projects/PROTNLM_EVALUATION/mammal-benchmark/paired-sequences/compare.py) using the two downloaded UniProt text records. See the [methods and limitations](../../../../projects/PROTNLM_EVALUATION/mammal-benchmark/paired-sequences/README.md).

- [x] Computed from sequence inputs; no hardcoded results.
- [x] The same script was run on ten different gene pairs.
- [x] Alignment completed and full output retained.
- [ ] Orthology, gene-model completeness and conservation of each biological claim require additional evidence. Sequence similarity alone does not establish them.

Current UniProt sequences were compared; their identity with the original ProtNLM input sequences has not been established.