You are on page 1of 15

BioMaxP : A Formal Approach for

Cellular Ion Pumps


Bogdan Aman

Gabriel Ciobanu

Romanian Academy, Institute of Computer Science


Bld. Carol I, no.8, 700505, Iasi, Rom
ania

FOI 2015
24-29 August 2015

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

1 / 15

Outline

Introduction

Syntax and Semantics of BioMaxP

Safety Automata

Relating BioMaxP to Safety Automata

Conclusion

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

2 / 15

Introduction

A fundamental mechanism in most of the living cells is the Na+ /K + ATPase that is essential for the maintenance of Na+ and K + concentrations across the membrane by transporting Na+ out of the cell and
K + back into the cell.
In this paper we model the movement of ions and the conformational transformations of ion transporters (NaK ion pumps, Na and
K ion channels) using a very simple but powerful new formalism called
BioMaxP .
BioMaxP allows to work with multisets of ions, explicit interpretation of
the transportation (from inside to outside, and from outside to inside)
based on the number of existing ions, and a maximal parallel execution
of the involved pumps.

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

3 / 15

Syntax and Semantics of BioMaxP


BioMaxP Syntax
Processes
P, Q ::=

amin !(v : T ) then P p


amax ?(f (u : T )) then P p
id (v ) p
P|Q

(sending)
(receiving)
(recursion)
(parallel)

A constraint min associated with a sending action amin !(z : T ) then P


makes the channel a available for sending z units/ions of type T only
if the total available quantity of ions of type T is greater than min.
A constraint max associated to a receiving action amax ?(x : T ) then P
along a channel a is activated only if the number of ions of the type T
available is less than max.

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

4 / 15

Syntax and Semantics of BioMaxP

BioMaxP Syntax
Processes
P, Q ::=

amin !(v : T ) then P p


amax ?(f (u : T )) then P p
id (v ) p
P|Q

(sending)
(receiving)
(recursion)
(parallel)

Remark
In order to focus on the local interaction aspects of BioMaxP , we abstract from arithmetical operations (using the function f ), considering by
default that the simple ones (comparing, addition, subtraction) are included
in the language.

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

5 / 15

Syntax and Semantics of BioMaxP

BioMaxP Operational Semantics


(Com)

v :T

min |T | max

and

{v /u}

amin !hv i then P | amax ?(f (u : T )) then P P | {v /u}P


and |T | = |T | v if f = id or |T | = |T | + v if f = add
id

{v /u}Pid
Pid

(Call)

id

id (v )

(Par1)

1
P1
P1

P 6

1
P1 | P
P1 | P

B.Aman; G.Ciobanu

Pid

def

where id (v : T ) = Pid

(Par2)

1
P1
P1

2
P2
P2

2
P1 | P2 1
P1 | P2

BioMaxP : A Formal Approach for Cellular Ion Pumps

6 / 15

Syntax and Semantics of BioMaxP


Example
Consider a system formed from n1 NaK pumps, n2 Na channels and
n3 K channels.
Each pump i is modelled by three processes:

one that models the interaction of the pump with the environment,
one modelling the interaction with the cell and
one that models the transport of ions through the membrane.

The molecular components are processes modelled as the ends of a


channel (one end for input, and another for output), while the molecular
interaction coincides with communication on channels.
NaKPumpEnv (id ) = site2[id ]160 ?(add(yna : NaEnv ))
then site2[id ]2 !h2K i
then p[id ]6 ?(add(yp : P))
then NaKPumpEnv (id )
B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

7 / 15

Safety Automata

Definition

A Safety Automata

An automaton A is a tuple hN, n0 , E i,


where
N is a finite set of nodes;

start
y<=20
10<=y
enter
x:=0, y:=0

n0 is the initial node;


E N B(C) NC N is
the set of edges.

x==10
work
x:=0

loop
y<=50

10<=y
y:=0

g ,a,r

n n is a shorthand notation for


hn, g , a, r , n i E . r denotes fresh assignments to variables after the transition is performed.

40<=y
leave
y:=0
end
y<=20

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

8 / 15

Safety Automata

Network of Safety Automata


Is the parallel composition A1 | . . . | An of a set of safety automata
A1 , . . . , An combined into a single system.
Synchronous communication inside the network is by handshake synchronisation of input and output actions.
In this case, the action alphabet consists of

a? symbols (for input actions),


a! symbols (for output actions),
symbols (for internal actions).

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

9 / 15

Safety Automata

A network state is a pair hn, ui, where n denotes a vector of current nodes
of the network (one for each automaton), and u is an assignment storing
the current values of all network integer variables.

Definition
The operational semantics of a automaton is a transition system where
states are pairs hn, ui and transitions are defined by the rules:
g ,,r

hn, ui
hn[ni /ni ], u i if ni ni , g |= u and u = r [u];

hn, ui
hn[ni /ni ][nj /nj ], u i if there exist i 6= j such that
1
2

gi ,a?,ri

gj ,a!,rj

ni ni , nj nj , gi gj |= u,
u = ri [rj [u]].

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

10 / 15

Relating BioMaxP to Safety Automata

Construction
Given a process P without the parallel operator at the top level, we associate
to it an automaton A = hN, n0 , E i, where n0 = l0 , N = {l0 }, E = . The
components N and E are updated depending on the structure of process P:
for P = amin !hv i then P1 we have

N = N {li +1} where i = max{j | lj N};


E = E {n, min |T |, a!, , li +1}.

for P = amax ?(f (u : T )) then P1 we have

N =(
N {li +1} where i = max{j | lj N};
E {li , |T | max, a!, |T | = |T | |u|, li +1 }, if f = id;
E=
E {li , |T | max, a!, |T | = |T | + |u|, li +1 }, if f = add.

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

11 / 15

Relating BioMaxP to Safety Automata

Building an automaton for each process leads to the next result about the
equivalence between a BioMaxP process P and its corresponding automaton
AP in state hnP , uP i (i.e., (AP , hnP , uP i). Their transition systems differ
not only in transitions, but also in states; thus, we adapt the notion of
bisimilarity:

Definition
A symmetric relation between BioMaxP processes and their corresponding

automata is a bisimulation if whenever (N, (AN , hnN , uN i)) if P


P ,

then hnP , uP i
hnP , uP i and (P , (AP , hnP , uP i)) for some P .

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

12 / 15

Relating BioMaxP to Safety Automata

Theorem
Given a BioMaxP process P, there exists an automata AP with a bisimilar
behaviour. Formally, P AP .

Corollary
For a BioMaxP process, the reachability problem is decidable.

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

13 / 15

Conclusion

In this paper we try to unify and extend our previous attempts to model
the movement of ions using the sodium-potassium pump by introducing
a simple, elegant and powerful new formalism called BioMaxP .
BioMaxP is able to:

capture the quantitative aspects (e.g., number of ions);


abstract conditions associated with evolution (e.g., the number of ions
is between certain thresholds).

This approach facilitates a better understanding of the processes happening in a cell viewed as a complex system of ion pumps working in
parallel.
As future work we plan to use Uppaal to verify some properties of the
systems modelled in BioMaxP .

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

14 / 15

Thank you!

B.Aman; G.Ciobanu

BioMaxP : A Formal Approach for Cellular Ion Pumps

15 / 15