convex rank tests
play

Convex rank tests Anne Shiu Texas A&M University CombinaTexas - PowerPoint PPT Presentation

Convex rank tests Anne Shiu Texas A&M University CombinaTexas 8 May 2016 From Algebraic Systems Biology: A Case Study for the Wnt Pathway (Elizabeth Gross, Heather Harrington, Zvi Rosen, Bernd Sturmfels 2016). Outline of talk


  1. Convex rank tests Anne Shiu Texas A&M University CombinaTexas 8 May 2016

  2. From Algebraic Systems Biology: A Case Study for the Wnt Pathway (Elizabeth Gross, Heather Harrington, Zvi Rosen, Bernd Sturmfels 2016).

  3. Outline of talk ◮ Introduction ◮ Main Result: convex rank tests = semigraphoids ◮ 2 counterexamples ◮ Application to biology

  4. Outline of talk ◮ Introduction ◮ Main Result: convex rank tests = semigraphoids ◮ 2 counterexamples ◮ Application to biology Joint work with Raymond Hemmecke, Jason Morton, Lior Pachter, Bernd Sturmfels, and Oliver Wienand.

  5. Introduction

  6. Preliminary definitions ◮ A fan in R n is a finite collection F of polyhedral cones such that: ◮ if C ∈ F and C ′ is a face of C , then C ′ ∈ F , and ◮ if C, C ′ ∈ F , then C ∩ C ′ is a face of C . ◮ The S n -arrangement (the braid arrangement ) is the arrangement of hyperplanes { x i = x j } in R n . ◮ Example: the fan associated to the S 3 - arrangement has 6 maximal cones. x 1 = x 3 x 1 = x 2

  7. What is a convex rank test? ◮ A rank test is a partition of S n . ◮ A convex rank test is a partition of S n defined by a fan that coarsens the S n -arrangement. ◮ Example: the following convex rank test partitions S 3 into 4 classes. 123 132 312 ( x 3 > x 1 > x 2 ) 213 231 321

  8. A non-convex rank test ◮ This partition of S 3 into 4 classes is not a convex rank test. 123 132 213 312 231 321 ◮ Remark : a convex rank test is determined by the walls removed from the S n -arrangement.

  9. Label walls by conditional-independence statements 123 132 1 ⊥ ⊥ 3 |∅ 213 312 1 ⊥ ⊥ 3 |{ 2 } 231 321 ◮ Two maximal cones of the S n -fan, labeled by permutations δ and δ ′ in S n , share a wall if δ and δ ′ differ by an adjacent transposition : there exists an index k such that δ k = δ ′ k +1 , δ k +1 = δ ′ k , and δ i = δ ′ i for i � = k, k + 1. ◮ Label wall { δ, δ ′ } by the conditional-independence (CI) statement: δ k ⊥ ⊥ δ k +1 | { δ 1 , . . . , δ k − 1 } .

  10. Conditional independence Consider a collection of n random variables indexed by [ n ] . [1 ⊥ ⊥ 2 |∅ ] [1 ⊥ ⊥ 3 |∅ ] [2 ⊥ ⊥ 3 |∅ ] [2 ⊥ ⊥ 3 | 1] [1 ⊥ ⊥ 3 | 2] [1 ⊥ ⊥ 2 | 3] [1 ⊥ ⊥ 4 |∅ ] [2 ⊥ ⊥ 4 |∅ ] [3 ⊥ ⊥ 4 |∅ ] [1 ⊥ ⊥ 2 | 4] [1 ⊥ ⊥ 3 | 4] [2 ⊥ ⊥ 3 | 4] [2 ⊥ ⊥ 4 | 1] [3 ⊥ ⊥ 4 | 1] [1 ⊥ ⊥ 4 | 2] [3 ⊥ ⊥ 4 | 2] [1 ⊥ ⊥ 4 | 3] [2 ⊥ ⊥ 4 | 3] [1 ⊥ ⊥ 2 | 34] [1 ⊥ ⊥ 3 | 24] [1 ⊥ ⊥ 4 | 23] [2 ⊥ ⊥ 3 | 14] [2 ⊥ ⊥ 4 | 13] [3 ⊥ ⊥ 4 | 12] [1 ⊥ ⊥ 5 |∅ ] [2 ⊥ ⊥ 5 |∅ ] . . . [4 ⊥ ⊥ 5 | 123] . . . The symbol [ i ⊥ ⊥ j | K ] represents the statement, “the random variables i and j are conditionally independent given the joint random variable K .”

  11. Semigraphoids ◮ (definition #1) A set M of CI statements on [ n ] is a semigraphoid if the following axiom holds 1 : ( SG ) If [ i ⊥ ⊥ j | K ∪ ℓ ] and [ i ⊥ ⊥ ℓ | K ] are in M then also [ i ⊥ ⊥ j | K ] and [ i ⊥ ⊥ ℓ | K ∪ j ] are in M . 1 Probabilistic Conditional Independence Structures , Studen´ y 2005

  12. Semigraphoids ◮ (definition #1) A set M of CI statements on [ n ] is a semigraphoid if the following axiom holds 1 : ( SG ) If [ i ⊥ ⊥ j | K ∪ ℓ ] and [ i ⊥ ⊥ ℓ | K ] are in M then also [ i ⊥ ⊥ j | K ] and [ i ⊥ ⊥ ℓ | K ∪ j ] are in M . ◮ Example: ( SG ) If [1 ⊥ ⊥ 2 | 3] and [1 ⊥ ⊥ 3 |∅ ] are in M , then also [1 ⊥ ⊥ 2 |∅ ] and [1 ⊥ ⊥ 3 | 2] are in M . ◮ So, M = { [1 ⊥ ⊥ 3 |∅ ] , [1 ⊥ ⊥ 2 | 3] } is not a semigraphoid. 123 132 1 ⊥ ⊥ 3 |{∅} 213 312 1 ⊥ ⊥ 2 |{ 3 } 231 321 1 Probabilistic Conditional Independence Structures , Studen´ y 2005

  13. Main result

  14. Main Theorem ◮ A convex rank test F is characterized by the collection of walls { δ, δ ′ } that are removed from the S n -arrangement. Let M F denote the CI statements that label those walls. ◮ Main theorem: The map F �→ M F is a bijection between convex rank tests and semigraphoids. ◮ The following convex rank test corresponds to the semigraphoid M = { 1 ⊥ ⊥ 3 |∅ , 1 ⊥ ⊥ 3 | 2 } . 123 132 1 ⊥ ⊥ 3 |∅ 213 312 1 ⊥ ⊥ 3 |{ 2 } 231 321

  15. Restating the Main result via the permutohedron

  16. The Permutohedron ◮ The fan of the S n -arrangement is the normal fan of the permutohedron P n (the convex hull of the vectors ( ρ 1 , . . . , ρ n ), where ρ is in S n ). 123 • 132 • 1 ⊥ ⊥ 3 |∅ 213 • • 312 1 ⊥ ⊥ 3 |{ 2 } 231 • • 321 ◮ The edges of the permutohedron correspond to walls of the S n -arrangement.

  17. The permutohedron P 4 2 ⊥ ⊥ 3 | 14 • • • • • • • • • • • • • 1432 • • 1423 4312 • 4213 ◦ 1342 • • 1243 4321 ◦ ◦ 4231 3412 • 2413 ◦ • • 3142 1234 • 2143 1324 • 3421 • ◦ 2431 2134 • 3124 • 3241 • ◦ 2341 • • 3214 2314 The 2-d faces of P n are squares and hexagons.

  18. Square and Hexagon Axioms Lemma: A set M of edges of the permutohedron P n is a semigraphoid if and only if M satisfies the following two axioms: ◮ Square axiom: Whenever an edge of a square is in M , then the opposite edge is also in M . ◮ Hexagon axiom: When two adjacent edges of a hexagon are in M , then the two opposite edges are also in M . 123 • 132 • 213 • • 312 231 • • 321 Main theorem, restated. Coarsenings of the S n -fan are equivalent to subsets of edges of P n that satisfy the Square and Hexagon axioms . Generalization to other Coxeter arrangemts. Coarsenings = subsets of edges with the polygon property . (Nathan Reading 2012).

  19. Hexagon axiom illustrated Consider M = { 1 ⊥ ⊥ 3 |∅ , 1 ⊥ ⊥ 2 |{ 3 }} (again). It is not a convex rank test, because it violates the Hexagon axiom: 123 • 132 • 1 ⊥ ⊥ 3 |∅ 213 • • 312 1 ⊥ ⊥ 2 |{ 3 } 231 • • 321

  20. Main theorem illustrated 4132 • 4123 • • • • • • • • • • • • 1432 • • 1423 • • • • • • • • • • • 4312 • 4213 ◦ • • • • 1342 • • • 1243 • • 4321 ◦ • • ◦ • 4231 • • 3412 • 2413 ◦ • • • • • • • • • • • 3142 • 1234 • • 2143 1324 • • • • • • • 3421 • ◦ • • 2431 • • 2134 • • 3124 • • 3241 • ◦ 2341 • • • • • • • • • • • • • • 3214 2314 f = (16 , 24 , 10)

  21. 2 counterexamples

  22. Semigraphoids: another definition ◮ Each CI statement defines a linear form in 2 n unknowns h I for I ⊆ [ n ]: [ i ⊥ ⊥ j | K ] �→ − h ijK + h iK + h jK − h K . ◮ Non-negativity of these linear forms defines the (2 n − n − 1)-dimensional submodular cone in R 2 n . ◮ The linear relations among the forms are spanned by entropy equations: [ i ⊥ ⊥ j | K ∪ ℓ ] + [ i ⊥ ⊥ ℓ | K ] = [ i ⊥ ⊥ j | K ] + [ i ⊥ ⊥ ℓ | K ∪ j ] .

  23. Semigraphoids: another definition ◮ Each CI statement defines a linear form in 2 n unknowns h I for I ⊆ [ n ]: [ i ⊥ ⊥ j | K ] �→ − h ijK + h iK + h jK − h K . ◮ Non-negativity of these linear forms defines the (2 n − n − 1)-dimensional submodular cone in R 2 n . ◮ The linear relations among the forms are spanned by entropy equations: [ i ⊥ ⊥ j | K ∪ ℓ ] + [ i ⊥ ⊥ ℓ | K ] = [ i ⊥ ⊥ j | K ] + [ i ⊥ ⊥ ℓ | K ∪ j ] . ◮ (definition #4) A semigraphoid M specifies the possible zeros for a non-negative solution of the entropy equations. ◮ A semigraphoid M is submodular if it is the set of actual zeros of a point in the submodular cone.

  24. Question 1 ◮ Postnikov, Reiner and Williams (2006) asked: Is every simplicial fan which coarsens the S n -fan the normal fan of convex polytope? ◮ Facts. A convex rank test F is the normal fan of a polytope if and only if the semigraphoid M F is submodular. This polytope is a generalized permutohedron . It is simple iff F is simplicial iff the posets on [ n ] are trees. ◮ The answer to the PRW question is no for n = 4: Proposition. This is simplicial, but not submodular: • • • • • • • • • • • • • • • ◦ • • • • • • ◦ ◦ • • • • ◦ • • • • • • • • • • • • ◦ • • • • • • • ◦ • • • • • • • •

  25. Proof: simplicial This simple polytope looks like a generalized permutohedron... 41 23 14 23 32 32 • • • • • • • • • • • • • • • 1342 4213 • ◦ ◦ ◦ ◦ ◦ 12 ◦ • 21 43 • ◦ • ◦ • ◦ • • ◦ • • ◦ 2413 1324 • ◦ ◦ ◦ ◦ • 4231 • • • • • 12 21 34 ◦ ◦ • ◦ ◦ 2431 ◦ ◦ • ◦ 23 14 41 M F = { [2 ⊥ ⊥ 3 | 14] , [1 ⊥ ⊥ 4 | 23] , [1 ⊥ ⊥ 2 |∅ ] , [3 ⊥ ⊥ 4 |∅ ] } .

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend