ACQ4: an open-source software platform for data acquisition and analysis in neurophysiology research

Abstract

The complexity of modern neurophysiology experiments requires specialized software to coordinate multiple acquisition devices and analyze the collected data. We have developed ACQ4, an open-source software platform for performing data acquisition and analysis in experimental neurophysiology. This software integrates the tasks of acquiring, managing, and analyzing experimental data. ACQ4 has been used primarily for standard patch-clamp electrophysiology, laser scanning photostimulation, multiphoton microscopy, intrinsic imaging, and calcium imaging. The system is highly modular, which facilitates the addition of new devices and functionality. The modules included with ACQ4 provide for rapid construction of acquisition protocols, live video display, and customizable analysis tools. Position-aware data collection allows automated construction of image mosaics and registration of images with 3-dimensional anatomical atlases. ACQ4 uses free and open-source tools including Python, NumPy/SciPy for numerical computation, PyQt for the user interface, and PyQtGraph for scientific graphics. Supported hardware includes cameras, patch clamp amplifiers, scanning mirrors, lasers, shutters, Pockels cells, motorized stages, and more. ACQ4 is available for download at http://www.acq4.org.

Keywords: analysis software; calcium imaging; data acquisition; electrophysiology; laser scanning photostimulation; multiphoton microscopy; patch clamp; python language.

Figure 1

Architecture of ACQ4. The central Manager is responsible for configuring devices and facilitating most communication between devices and modules. The user interface is composed of multiple modules, each providing a particular functionality. The Data Manager handles storage, the organization of raw data, and storage of associated metadata.

Figure 2

An example hierarchy of optomechanical devices. The optomechanical device arrangement allows the software to map specific locations on a sample to a pixel on the CCD, or to the proper pairs of scan mirror voltages. The position of the stage, objective magnification, and per-objective offset are all handled automatically. New devices may be added anywhere in the hierarchy, which allows ACQ4 to support arbitrary hardware configurations.

Figure 3

Screen capture of the Camera module during a mouse auditory cortex brain slice experiment. All experimental procedures were approved by the Institutional Animal Care and Use Committee at the University of North Carolina at Chapel Hill. The acquisition system consists of an NI-6259 data acquisition board, a MultiClamp 700A amplifier, a Photometrics QuantEM 512SC camera, and a custom multiphoton imaging system (Cambridge Technologies 6510H galvanometric scan mirrors, Coherent Chameleon Vision II ultrafast Ti:Sapphire laser, and a Hamamatsu H7422P-40 photomultiplier incorporated onto a modified Zeiss Axioskop 2 FS microscope with 5× and 63× objectives). The recording chamber was mounted on a 3-axis motorized stage (Mike's Machine Co., Boston, MA) driven by a Sutter MP-285 controller (however, use of the MPC-200 instead of the MP-285 is strongly preferred). (A) 2-photon image of fluorescence from an Alexa Fluor 568 labeled cortical neuron recorded in whole-cell tight seal mode. The red rectangle is a draggable region defining the area to be imaged. (B) Live video from CCD camera. (C) Background frames previously acquired with camera provided a wide-field, persistent view of the brain slice.

Figure 4

Screen capture of the TaskRunner module running a calcium imaging task during recordings from an mouse auditory cortex neuron in a brain slice experiment. The acquisition system is described in Figure 3. The cortical neuron is filled with the calcium indicator Fluo-4 (200 micromolar) and is electrically stimulated through the patch pipette. Additional control panels for selecting devices, running protocol sequences, and configuring the data acquisition board are hidden. To design this task, the experimenter has already selected the camera and patch-clamp channel to be included, and has rearranged the panels to optimize use of the window space. The task has been executed once, and the results are displayed in the rightmost panels. (A) Control panel for configuring the behavior of the patch clamp amplifier, including the output waveform specification. (B) Plot showing the most recent electrode recording and the command waveform. (C) Interface for controlling the Camera. This includes control over the camera's frame transfer mode and triggering waveform. (D) The recorded video data is displayed and a region of interest defines the pixels that are averaged together and plotted in the traces below, showing the calcium transient evoked by action potentials in the cell. The bottom-most plot shows the exposure times of acquired camera frames for reference to the electrical recording.

Figure 5

Screen capture of the Patch module after patching a neuron in a mouse cochlear nucleus brain slice. In this image, approximately 40 s have passed since the membrane was ruptured to begin whole-cell recording. This data was acquired using only an NI-6259 acquisition board and a MultiClamp 700A amplifier. (A) Plots showing the voltage clamp recording and command waveform. This data is used by the module to determine the membrane properties shown at center-left (input resistance, access resistance, etc.). (B) Plot showing history of input resistance over the last 80 s. Maximum seal resistance and time of break-in are both recorded in this data.

Figure 6

Screen capture of the Mosaic Editor analysis module showing reconstruction and atlas registration of a mouse cochlear nucleus brain slice imaged under a 5× microscope objective. Although the Mosaic Editor is most commonly used to simply reconstruct image mosaics, the analysis shown here is a step in registration of the data taken during the experiment to a 3D atlas of the nucleus. The images were collected with a Photometrics Quantix-57 camera and Zeiss Axioskop 2 FS microscope mounted on a manual translation stage. Positioning information was generated from a custom set of rotary optical encoders driven by the stage micrometers. (A) Volumetric rendering of a 3D atlas of the mouse cochlear nucleus. The white rectangle indicates the region of the nucleus from which the brain slice was taken, as determined by photos taken during the slicing procedure. This is used to create a digital slice of the 3D atlas, which is overlaid on (B) an automatically-reconstructed mosaic of several tiled photos of the brain slice.

Figure 7

Screen capture of the Photostim analysis module, processing laser-scanning photostimulation mapping using glutamate uncaging (300 μM MNI-glutamate) data from a mouse cochlear nucleus brain slice experiment. The hardware used to collect the data for this experiment is described in Figure 3. A Q-switched 355 nm DPSS laser was projected through scan mirrors to uncage glutamate at locations on the slice indicated by the displayed map results. (A) Analyzed results from a photostimulation map overlaid on images of the brain slice used in this experiment. Brightly colored circles indicate that a strong synaptic response was detected in the patched cell when the spot was photostimulated, whereas transparent circles indicate that no response was detected. (B) Plot of data from a single photostimulation recording. The traces in blue indicate the amplitude and time course of the evoked synaptic currents. (C) A diagnostic plot showing the same data at an intermediate filtering stage in which the onset of synaptic events has been detected.