K9IWH5 Research Notes

Key findings

• UniProt names this protein lysozyme [file:DESRO/K9IWH5/K9IWH5-uniprot.txt "RecName: Full=lysozyme"].
• UniProt assigns this protein to glycosyl hydrolase family 22 [file:DESRO/K9IWH5/K9IWH5-uniprot.txt "Belongs to the glycosyl hydrolase 22 family."].
• Deep research identifies LYZ as a c-type lysozyme in the GH22 family [file:DESRO/K9IWH5/K9IWH5-deep-research-falcon.md "LYZ corresponds to a c-type lysozyme (muramidase; EC 3.2.1.17) that hydrolyzes peptidoglycan and belongs to the classical GH22 lysozyme family."].

2026-07-31 compliance review

There is a gene-specific paper, and it was not being used. The UniProt entry
lists three EMBL records; the third (MK561737 / QEQ43371.1, liver) carries
RX PubMed=31506780 — He et al. 2019, Adaptive Evolution of C-Type Lysozyme in
Vampire Bats. Fetched into the cache (abstract only; full_text_available: false). Key points:

• PMID:31506780 — no paralogue in vampire bats, unlike the insectivorous bat clade
  that duplicated lysozyme for chitin digestion.
• PMID:31506780 and PMID:31506780.
• The authors' reading — PMID:31506780 — is an
  interpretation of a selection analysis, not a measurement. Tagged as such in
  finding_review and turned into a knowledge gap rather than an annotation.

The second salivary-gland EMBL record (JAA45151.1) traces to the Vampirome study
(PMID:23411029), which puts lysozyme in the accessory gland at both transcript
and protein level PMID:23411029 and detects it directly by LC-MS/MS PMID:23411029.

Reversed three annotation calls made by the previous version. All three were
in the direction of throwing information away:

• GO:0050830 (Gram-positive) and GO:0050829 (Gram-negative) had been set to
  MODIFY → GO:0042742 defense response to bacterium. That generalises two
  specific, mechanistically explained terms into a parent that is already
  separately annotated and accepted in the same file — pure information loss.
  Both are now ACCEPT. Cross-checked against human LYZ (P61626) in QuickGO:
  GO:0050830 is IDA (PMID:21093056), GO:0050829 is IBA — the same TreeGrafter
  propagation applied here.
• GO:0031640 killing of cells of another organism had been marked
  over-annotated for being "too broad". It is not an inflation: bacteriolysis is
  the direct outcome of muramidase action, the keyword source is literally
  "Bacteriolytic enzyme", and human LYZ carries the term by IDA
  (PMID:9727055). Now ACCEPT.

Filled a missing GO aspect. The GOA record has no cellular component term at
all. Added GO:0005576 extracellular region as a NEW (ISS) annotation, supported
independently by the signal peptide (SIGNAL 1..18), by the human orthologue's
IDA annotation to the same term, and by proteomic detection in the bat gland.

gene_symbol changed from the accession to LYZ, matching both the UniProt
GN Name=LYZ line and the GOA SYMBOL column.