The feature extraction GUI allows users to select data from an
HBP dataset and/or upload their own, and extract
electrophysiological features of interest.
The application leverages the Python Electrophys Feature Extract Library (eFEL) and provides a friendly interface to select both individual voltage traces (based on the stimulus current applied) and features to be extracted.
Step 1 of 4Select the data from a dataset, based on cell properties to be chosen from the filter dropdown menus. Additionally and/or alternatively upload your own data for processing (see Fig. 1).
Step 2 of 4Once the filtering/upload done, individual traces are plotted and can be checked for selection. Traces represent the voltage membrane responses of the selected cell to the stimulus displayed in the legend. Highlight individual traces by hovering on the corresponding stimulus and select them by clicking on the same legend or through the selection buttons (see Fig. 2). When all the traces of interest have been selected, go the next page for feature selection.
Step 3 of 4The feature selection page allows you to select the features to extract from the electrophysiological chosen signals. Features are grouped by type: 1) Spike event features - 2) Spike shape features and 3) Voltage features, and are selected by clicking on the corresponding box. When hovering on feature names a brief description is provided (see Fig. 3).
Step 4 of 4The feature extraction process computes the means and standard deviations of the selected features for individual cells and for individual stimuli. Once a stimulus is chosen for a single file, it will be taken into account for all the selected files corresponding to the same cell. Finally, a grand mean is computed among different cells.