Suppose we want to minimize the Euclidean distance of the solution to the origin while subject to linear constraints. This will require a quadratic objective function. Consider this example problem:

min $$\frac{1}{2}\left(x^2 + y^2 \right)$$

subject to

$$x + y = 2$$

$$x \ge 0$$

$$y \ge 0$$

This problem can be visualized graphically:

In :

%matplotlib inline
import plot_helper

plot_helper.plot_qp1() The objective can be rewritten as $$\frac{1}{2} v^T \cdot \mathbf Q \cdot v$$, where $$v = \left(\begin{matrix} x \\ y\end{matrix} \right)$$ and $$\mathbf Q = \left(\begin{matrix} 1 & 0\\ 0 & 1 \end{matrix}\right)$$

The matrix $$\mathbf Q$$ can be passed into a cobra model as the quadratic objective.

In :

import scipy

from cobra import Reaction, Metabolite, Model, solvers


The quadratic objective $$\mathbf Q$$ should be formatted as a scipy sparse matrix.

In :

Q = scipy.sparse.eye(2).todok()
Q

Out:

<2x2 sparse matrix of type '<class 'numpy.float64'>'
with 2 stored elements in Dictionary Of Keys format>


In this case, the quadratic objective is simply the identity matrix

In :

Q.todense()

Out:

matrix([[ 1.,  0.],
[ 0.,  1.]])


We need to use a solver that supports quadratic programming, such as gurobi or cplex. If a solver which supports quadratic programming is installed, this function will return its name.

In :

print(solvers.get_solver_name(qp=True))

cplex

In :

c = Metabolite("c")
c._bound = 2
x = Reaction("x")
y = Reaction("y")
m = Model()
sol.x_dict

Out:

{'x': 1.0, 'y': 1.0}


Suppose we change the problem to have a mixed linear and quadratic objective.

min $$\frac{1}{2}\left(x^2 + y^2 \right) - y$$

subject to

$$x + y = 2$$

$$x \ge 0$$

$$y \ge 0$$

Graphically, this would be

In :

plot_helper.plot_qp2() QP solvers in cobrapy will combine linear and quadratic coefficients. The linear portion will be obtained from the same objective_coefficient attribute used with LP’s.

In :

y.objective_coefficient = -1

Out:

{'x': 0.5, 'y': 1.5}