goFRAME

goFRAME

a Go-based Framework for Rapid Analytical Modeling in Electrophysiology (FRAME). goFRAME is a simple directory that can greatly improve the efficiency and organization of electrophysiological model simulations.

Authors

• @mylesmax

Working Examples

/examples/hodgkinhuxley

original paper : https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1392413/

/examples/luorudy1991

original paper : https://pubmed.ncbi.nlm.nih.gov/1709839/

Usage

Download

Use the following command to download the repository:

go get -t github.com/mylesmax/goFRAME

Model Generation

It is recommended to read the steps below for a general understanding of the framework, but for a more streamlined experience, follow the template tutorial in /template.

Parameter, gate, and current formulation is demonstrated in detail in the examples and template. However, here are some key notes:

Parameters should be defined in the initial State, which is recommended to be generated in a func init(). The general form of a State is:

type State struct {
	V, Cm, RTF, Stim, T float64
	Index                int
	//Nernst Potentials
	E map[string]float64

	//conductances
	GBar map[string]float64

	//currents
	I map[string]float64

	//gates
	Gate map[string]float64

	//ion concentrations
	ConcOut map[string]float64
	ConcIn map[string]float64

	//misc
	Misc map[string]float64
}

For the initial State, set V=Vrest and Index=0.

If a gate is defined by alpha and beta values (GateAB), it should be defined as

goFRAME.GateAB{Alpha, Beta, ID}

where Alpha and Beta are func (V float64) float64 and ID is of type string.

If a gate is defined by an analytical/direct solution (GateDirect), it should be defined as

goFRAME.GateDirect{Ss, ID}

where Ss is func(V float64) float64.

In both cases, the ID field should be matched to the gate name (e.g., ID is "m" for gate m).

The params...interface{} is flexible to three parameters: params[0] is the current State, params[1] is an Out (i.e., all saved States, which is []State), and params[2] is an int that represents the index.

Most models have similar calculations of dvdt, the change in membrane voltage, which is given in examples. For further customization, the params...interface{} is flexible to three parameters: params[0] is the current Solver, params[1] is an int that represents the index, and params[2] is a float64 that represents the sum of all currents.

Define a stimulus array Stims that contains as many Stim instances as needed:

type Stim struct{
	Start, End, Intensity float64
}

Define a TSet array that contains the simulation start time, simulation end time, and dt:

type TSet struct {
	Start, Dt, End float64
}

Then, generate an array to capture the evolution of the system, and save its initial value to be the initial State. For example:

out = []goFRAME.State{initial}

Now, define the solver:

type Solver struct {
	Out         Out
	Tset        TSet
	Stims       Stims
	GatesAB     []GateAB
	GatesDirect []GateDirect
	Currents    []Current
	ToIntegrate map[string]func(params ...interface{}) float64
	Method      string
}

For usable methods, see the bottom of the README. After the Solver has been generated, you may solve the system with

solver.Solve()

and write to excel with

WriteExcel(O Out, filePath string)

Usable Numerical Methods

"RK1" : Euler Method (Runge-Kutta 1)

License