Page tree
Skip to end of metadata
Go to start of metadata

Goals

  • Connect translocation terms to distinct cellular pools of an abundance (input pool and output pool)
    • In BEL v1.0, sub-cellular location can only be specified via statement annotations
  • Improve clarity and semantic checking of translocation expressions
    • no BEL function explicitly attached to locations within translocation terms
    • to and from locations inferred by position in term
    • location values have encoding 'A' (abundance) which is too general to facilitate semantic checks

BEL v1.0 translocation syntax:

tloc(p(HGNC:AKT1), MESHCL:Cytoplasm, MESHCL:"Cell Membrane Structures")

In BEL v1.0, the tloc() function takes an abundance term and two locations. These locations are generally from the MESH Cellular Structure or GO cellular component namespaces, and should have encoding 'A' (abundance). The above example term represents translocation of human AKT1 protein from the cytoplasm to cell membranes.

Proposal:

  • introduce location() as inner function (similar to a protein modification or variant)
    • provide new encoding type for namespace values that can be used as locations
  • modify tloc() expression by clearly designating "from location" and "to location" functions for inner location terms

Examples:

Translocation of AKT1 protein from cytoplasm to membranes:

tloc(p(HGNC:AKT1), fromLoc(MESHCL:Cytoplasm), toLoc(MESHCL:"Cell Membrane Structures"))

Designation of cytoplasmic pool of AKT1 protein:

p(HGNC:AKT1, loc(MESHCL:Cytoplasm))

loc() modification should apply to most/all abundance types - a(), g(), r(), m(), p(), complex().

Compiler Expansions:

Compiler expansions will require introduction of compiler-introduced relationships. These examples provide an overview of network connectivity.

Proposed compiler expansions - inferred from translocation term

tloc(p(HGNC:AKT1), fromLoc(MESHCL:Cytoplasm), toLoc(MESHCL:"Cell Membrane Structures")) output p(HGNC:AKT1, loc(MESHCL:"Cell Membrane Structures"))

p(HGNC:AKT1, loc(MESHCL:"Cytoplasm")) input tloc(p(HGNC:AKT1), fromLoc(MESHCL:Cytoplasm), toLoc(MESHCL:"Cell Membrane Structures")) 

The translocation of AKT1 protein from cytoplasm to cell membranes, is treated like a reaction, with cytoplasmic AKT1 treated as the input/reactant and membrane associated AKT1 as the output/product.

Proposed compiler expansions - inferred from sec() translocation term:

sec(p(HGNC:TNF)) output p(HGNC:TNF, loc(MESHCL:"Extracellular Space"))

p(HGNC:TNF, loc(MESHCL:"Intracellular Space")) input sec(p(HGNC:TNF)) 

Proposed compiler expansions - inferred from loc term

p(HGNC:AKT1) hasSubset p(HGNC:AKT1, loc(MESHCL:Cytoplasm))

Notes

  • Are statement annotations for cellular location still useful given the ability to add locations for all abundance terms?
  • Can loc(), sub(), and pmod() be used within the same abundance term?
  • Should loc() be used within activity terms, e.g., kin(p(HGNC:AKT1, loc(MESHCL:Cytoplasm))
    • If so, should hasSubset extend to connecting activities?
  • We need to think about ways to propagate causality such that we are not changing the size of the total available pool of something by increasing/decreasing its modification or changing its location.
  • No labels