Skip to content

Depletion edges are emitted before boundary expansion and suppress assembly for the complexes they target #68

Description

@adamjohnwright

_emit_substrate_depletion_edges is called at src/logic_network_generator.py:1925; _emit_boundary_decomposition_edges at :1937. The latter computes root_uuids = sources - targets.

A depletion edge adds catalyst -> input, which makes that input a target. So a root-input complex that a depletion edge points at stops being a root and loses its assembly expansion — its subunits never get member->complex edges, so the individual proteins are no longer addressable for perturbation at that node.

Verified: 13 root-input complex occurrences catalog-wide lose their assembly expansion this way — Signaling_by_PDGF 3, Signaling_by_EGFR 2, Semaphorin 2, ROBO 2, and 4 singletons.

Small, but it is an emission-order artifact rather than a modelling decision: whether a complex's subunits are addressable should not depend on whether a phosphatase happens to deplete it. Computing root_uuids before the depletion pass (or excluding depletion edges from the targets set) would fix it.

Worth a DeltaSignal before/after like any connectivity change, though the expected effect is small — the perturbed-gene coverage of 13 complexes across the catalog.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions