ERG1 notes

Workflow

• The falcon/perplexity-lite deep-research recipe was attempted before this manual pass and failed because
  this environment has no agentapi command and no provider API key. No
  ERG1-deep-research-falcon.md or ERG1-deep-research-perplexity-lite.md file was created; these notes
  summarize the cached literature and family rows instead.
• I searched PubMed/Web for newer Candida albicans ERG1 or squalene epoxidase papers. Recent hits were
  compound-perturbation, docking, or adjuvant studies that did not change the direct ERG1 functional
  model, so no new PMID cache files were needed.

Manual evidence notes

• UniProt Q92206 records Candida ERG1 as the PTHR10835 squalene monooxygenase and gives the curated
  Rhea reaction for squalene epoxidation: "Reaction=squalene + reduced [NADPH--hemoprotein reductase]
• O2 = (S)-" [file:genes/CANAL/ERG1/ERG1-uniprot.txt].
• Ryder and Dupont assayed the Candida particulate enzyme before ERG1 was cloned: "A washed
  'microsomal' fraction converted radiolabelled squalene to 2,3-oxidosqualene and lanosterol" and its
  minimum activity requirements were oxygen, NADH or NADPH, and FAD PMID:6378256.
• Ryder and Dupont then showed that allylamines inhibit the Candida enzyme; naftifine and SF 86-327 were
  "potent inhibitors of the Candida epoxidases" PMID:3877503. This supports druggability of the enzyme
  but the molecular assignment to ERG1 comes from the later cloning paper.
• Favre and Ryder cloned Candida squalene epoxidase and expressed it in budding yeast; homogenates
  "overexpressing the C. albicans protein SE activity was 10-fold higher than the" endogenous activity
  PMID:9161422.
• Pasrija et al. conditionally depleted ERG1 and found "a total lack of ergosterol and instead
  accumulation of squalene" PMID:15845783. They also reported drug-sensitivity and hyphal defects, but
  those are downstream consequences of losing ergosterol-rich membrane structure rather than direct ERG1
  processes.
• The later raft-lipid report uses the same logic: ERG1 or IPT1 perturbation "leads to improper surface
  localization of" Cdr1p PMID:16246085. This is useful phenotype evidence for sterol/sphingolipid-rich
  domains, not a reason to make Cdr1 localization a core ERG1 annotation.
• Roemer et al. placed one ERG1 allele under MET3p and showed the conditional strain was specifically
  hypersensitive to terbinafine: "reduced expression of Ca ERG1 resulted in cells that were hypersensitive
  only to the target-specific inhibitor terbinafine" PMID:16607011.
• The CaFT paper reports genome-wide heterozygote responses to many compounds, with "additional inhibitors
  of ergosterol biosynthesis, including terbinafine, lovastatin, and dyclonine" in the supplements
  PMID:17604452. The xenobiotic-response GO rows inferred from these profiles are target/fitness
  perturbations, not evidence that ERG1 executes xenobiotic response.
• The plasma-membrane proteome paper identified 214 proteins by LC-MS/MS PMID:19824013, but the local
  cache contains only the abstract and does not expose the ERG1-specific entry. I left the plasma-membrane
  row undecided pending the full protein table.

PANTHER rows

• PTHR10835 PTN000089417 GO:0004506 F IBD false supports fungal/animal squalene
  monooxygenase activity.
• PTHR10835 PTN000089417 GO:0005783 C IBD false supports ER localization for the same
  broad squalene monooxygenase node.
• PTHR10835 PTN000851137 GO:0006696 P IBD false supports the fungal ergosterol biosynthetic
  process assertion.