CG18507 vs TMEM268: isoform and ortholog comparison

Script: compare_tmem268.py (run with uv run compare_tmem268.py; sequences fetched live from UniProt REST).
Raw output: results-tmem268.json.

CG18507 UniProt products (FBgn0028527)

All five UniProt entries for CG18507 are unreviewed TrEMBL records; there is no Swiss-Prot canonical entry.

Accession	Length	Residues identical to M9PBB3 from the N-terminus
M9PBB3	510	510 (benchmark accession)
M9PCX0	510	491
M9PFU6	499	492
Q9VJW7	513	492
Q7KTB0	514	438

The five products share an identical N-terminal region, including both Phobius transmembrane helices (292-310, 316-339)
and the region that aligns to vertebrate TMEM268. They differ only in their C-terminal 20-75 residues, consistent with
alternative 3' exons. M9PBB3 is a full-length alternative product of the gene, not a fragment.

M9PBB3 vs vertebrate TMEM268 (same PANTHER subfamily PTHR31193:SF1)

Local alignment, BLOSUM62, gap open -11 / extend -1:

Target	Fly span	Target span	Aligned pairs	Identity	Similarity
Human TMEM268 Q5VZI3 (342 aa)	227-399	39-215	164	30.5%	52.4%
Mouse Tmem268 Q8R239 (342 aa)	227-486	40-331	251	29.1%	49.0%

The aligned core spans the two predicted transmembrane helices of both proteins (fly 292-339; human UniProt TRANSMEM 105-125
and 132-152). Fly residues 1-226 (largely disordered and low-complexity per MobiDB-lite) have no counterpart in the
vertebrate proteins, so the fly protein is an N-terminally extended member of the family.

Interpretation and limits

Pairwise similarity is consistent with, but does not by itself establish, orthology; the ortholog assignment rests on
shared PANTHER subfamily membership (PTHR31193:SF1, a single-subfamily family).
This analysis says nothing about whether the integrin-stabilizing role reported for mammalian TMEM268 is conserved;
Drosophila has no beta4 integrin or CD11b ortholog.