G1/S transition of the mitotic cell cycle (Start / restriction point)

A taxon-neutral decomposition of the G1/S transition: the commitment step ("Start" in yeasts, the "restriction point" in metazoa) at which a cell irreversibly enters a new division cycle. Four steps are modeled: (1) a G1 cyclin-CDK (cyclin D-CDK4/6 in metazoa, Cln3-Cdc28 in budding yeast, Puc1/Cig-Cdc2 in fission yeast) accumulates with growth and phosphorylates the transcriptional corepressor of the G1/S program; (2) the G1/S transcription factor is released from its corepressor (Rb from E2F-DP in metazoa, Whi5 from SBF/MBF in budding yeast; fission yeast MBF has no Rb/Whi5-type corepressor and is modeled as its own variant); (3) the G1/S cyclins transcribed by that program (cyclin E, Cln1/2 and Clb5/6, Cig2) activate the G1/S CDK; and (4) the G1/S CDK phosphorylates the stoichiometric CDK inhibitor (p27/Kip1, Sic1, Rum1), marking it for SCF-dependent ubiquitination and proteasomal destruction, which releases S-phase CDK activity in a switch-like positive feedback loop. The module ends with active S-phase cyclin-CDK; DNA replication licensing and initiation are downstream and are not part of this module. Representative members are drawn from human, Saccharomyces cerevisiae and Schizosaccharomyces pombe. Grounded in GO:0000082 (G1/S transition of mitotic cell cycle).

Boundary decisions. The module starts with growth-coupled accumulation of the G1 cyclin-CDK and ends with active S-phase cyclin-CDK after CKI destruction; origin licensing (MCM loading) and replication initiation (CDC7/DBF4, Cdc45/GINS) are separate modules. The DNA-damage G1 checkpoint (p53/p21, Rad53/Mec1) and mitogen signalling into cyclin D are regulatory context, not parts. Only the corepressor-release step is split by taxon: fission yeast has no Rb/Whi5-type corepressor, so its MBF (Cdc10-Res1-Res2-Rep2) is modeled as a third variant whose CDK-dependent trigger is a knowledge gap (Nrm1-mediated negative feedback and Cdc2-Cig2 regulation are documented but not modeled here as parts). The G1/S cyclin annoton deliberately spans two PANTHER families (PTHR10177 cyclins and PTHR21615 Cln1/Cln2) because the budding-yeast Start cyclins Cln1/2 are not classified with cyclin E, Clb5/6 or Cig2. The CKI annoton carries no PANTHER family term: p27/Kip1 belongs to PTHR10265 whereas Sic1 and Rum1 are unclassified and structurally unrelated to Cip/Kip proteins, so the family descriptor grounds only through representative members and an ANY_WITH_FUNCTION-like description. The F-box adaptor unit of SCF is likewise split across families (Skp2, Cdc4, Pop1, Pop2) and declared with family_terms using the local PANTHER 19.0 classification, which disagrees with the UniProt PANTHER cross-references for these four proteins. Deep research: no module deep-research report accompanies this file (no provider credentials were available in the authoring session; writing one by hand under a provider name is prohibited). Open questions: whether the fission-yeast Start trigger is best modeled as Cdc2-Cig2 phosphorylation of MBF components or of Rum1; and the extent to which metazoan cyclin E-CDK2 versus cyclin D-CDK4/6 is rate-limiting for Rb inactivation in vivo.