Search

We construct neuron models from data by transferring information from an observed time series to the state variables and parameters of Hodgkin-Huxley models. When the learning period completes, the model will predict additional observations and its parameters uniquely characterise the complement of ion channels. However, the assimilation of biological data, as opposed to model data, is complicated by the lack of knowledge of the true neuron equations. Reliance on guessed conductance models is plagued with multi-valued parameter solutions. Here, we report on the distributions of parameters and currents predicted with intentionally erroneous models, over-specified models, and an approximate model fitting hippocampal neuron data. We introduce a recursive piecewise data assimilation (RPDA) algorithm that converges with near-perfect reliability when the model is known. When the model is unknown, we show model error introduces correlations between certain parameters. The ionic currents reconstructed from these parameters are excellent predictors of true currents and carry a higher degree of confidence, >95.5%, than underlying parameters, >53%. Unexpressed ionic currents are correctly filtered out even in the presence of mild model error. When the model is unknown, the covariance eigenvalues of parameter estimates are found to be a good gauge of model error. Our results suggest that biological information may be retrieved from data by focussing on current estimates rather than parameters.

Many neurological diseases originate in the dysfunction of cellular ion channels. Their diagnosis presents a challenge especially when alterations in the complement of ion channels are a priori unknown. Current approaches based on voltage clamps lack the throughput necessary to identify the mutations causing changes in electrical activity. Here, we introduce a single-shot method for diagnosing changes in the complement of ion channels from changes in the electrical activity of a cell. We developed data assimilation (DA) to estimate the parameters of individual ion channels and from these parameters reconstruct the ionic currents of hippocampal CA1 neurons to within 11% of their actual value. DA correctly predicts which ionic current is altered and by how much after we blocked the BK, SK, A and HCN channels with selective antagonists of known potency. We anticipate our assay will transform the treatment of neurological disease through comprehensive diagnosis and drug screening., Comment: 42 pages, 6 figures, 3 tables