Welcome to DISTING [1], a tool for generating alternative structurally identifiable linear compartmental (LC) models* that are input-output indistinguishable from a postulated LC model [2,3].

App Usage
The site is centered around a job queue showing job status and results. Users can create, rerun or check job status in the queue.

User Entry of Model to be Tested (Model 0): This is done on the add job page simply by checking boxes where entries exist in one matrix and 2 vectors. The first is the adjacency (connection) matrix A of graph theory, where entry ai,j is checked to indicate a link to compartment i from compartment j; and ai,i is checked to indicate elimination (a leak) from compartment i. Vectors R and M get checked entries to specify the input and output locations in the LC model.

Graphical results show the original Model 0 and the structurally identifiable Models 1, 2, …, all downloadable in png format using Google Chrome, Firefox or Safari. The models can be interactively rearranged by left-clicking and dragging on a node.

Jobs start in pending status, move to running (one at a time) and finally to complete If an error occurs, the status changes to error. Job processing time depends on the number of simultaneous server job requests.

* LC models can be expressed by a set of first-order differential equations in vector-matrix form: dx/dt = Kx + Bu with outputs: y = Cx. Here x are state variables (quantities or concentrations). The compartment matrix K consists of rate constants between compartments (off-diagonal); and of the negative of the sum of rate constants from each compartment, including leaks to the environment (on-diagonal). The input vector B contains input gains, while the output vector C contains measurement gains. DISTING requires only the connectivity of LC models to run, expressed in the equivalent model graph theory representation A, R, M.

  1. Davidson, N.R., Godfrey, K.R., Alquaddoomi, F., Nola, D. and DiStefano III, J.J., 2017. DISTING: A web application for fast algorithmic computation of alternative indistinguishable linear compartmental models. Computer methods and programs in biomedicine, 143, pp.129-135.
  2. M.J. Chapman and K.R. Godfrey (1989): A methodology for compartmental model indistinguishability, Math Biosci, 96, pp. 141-164
  3. L-Q. Zhang, J.C. Collins and P.H. King (1991): Indistinguishability and identifiability analysis of linear compartmental models, Math Biosci, 103, pp.77-95

Actions