gPRIME Data Acquisition

1
gPRIME Data Acquisition Analysis Software for
Physiology in Research and Education Gus K. Lott
III1,2, Bruce R. Johnson1, Robert H. Bonow1,
Bruce R. Land1, and Ronald R. Hoy11Department of
Neurobiology and Behavior, Cornell Univ., Ithaca,
NY2Howard Hughes Medical Institute Janelia
Farm Research Campus
Introduction This software project grew out of a
confluence of events at Cornell University in the
Spring of 2007. The undergraduate
neurophysiology lab course at Cornell needed an
acquisition and analysis tech refresh and a PhD
student biophysical electrical engineer was
brought on as a teaching assistant. The result
was gPRIME (Physiology Recording Identification
of Multiple Events). This software package was
developed week to week in the course as students
moved from lab exercise to lab exercise. A new
version of the software would be developed and
deployed and the students would carry out the
weeks exercise using new features and provide
feedback about usability and bugs. This rapid
design and feedback process was made possible by
the Matlab (Mathworks Inc.) development
environment. Matlabs powerful engineering tools
allow for rapid code development and testing
while the Matlab Compiler toolbox allows for any
code to be exported to a stand-alone executable
file for deployment. As the software developed,
researchers in the Neurobiology Behavior
department at Cornell began to take an interest.
The tool transferred to active research projects
and feedback from both researchers and students
drove development. At the end of the spring
semester, student projects in the physiology lab
course where reproducing and exceeding published
work from recent articles from a variety of well
funded groups. For example, students were
characterizing crayfish neuromuscular junction
response to a variety of chemicals (including DDT
and caffeine), and were characterizing nerve
conduction velocity as a function of temperature
in the crayfish. Other students explored network
connectivity and synchrony in multiple nerves of
the crayfish tail. Since its development, gPRIME
has expanded to applications in multiple
university level courses and enables global
collaborative projects in high schools while
still acting as a work-horse for several
distinguished researchers around the world from
behavioral physiology in invertebrates to deep
brain recordings in mammalian systems.
Developed for students researchers in a
collaborative feedback loop
Data Grooming and Analysis of Clusters Clusters
of spikes may be extracted from a 2 dimensional
space using an arbitrary polygon or box.
Resulting spike clusters may be groomed (click on
an outlier and hit the delete key, see right) to
make sure that your results accurately represent
the data or simply to visualize where a certain
event is located in the data space. This mode
also allows for visualization of the rate
behavior of a specific subset of events and the
location of the events in the raw data channel.
Real Time and Offline Correlations gPRIME allows
the user to correlate a set of time points with
any source in both real time and offline modes.
The students auto-correlate a spike in real time
to produce SNR enhancement of an otherwise noisy
spike (top). In the bottom image, students have
cross-correlated a spike in an innervating neuron
with the post synaptic potential produced by that
neuron in the muscle. This feature is a powerful
tool to extract time correlated signals from
otherwise uncorrelated background noise in any
experiment where a triggering signal is available.
Current Availability gPRIME is currently
available from a Cornell University based web
site associated with the neurophysiology lab
course for which it was initially
developed. http//crawdad.cornell.edu/gprime/ The
software consists of a 180MB download and the
web site contains documentation, application
examples, real data from experiments, and
step-by-step instructions on getting started with
your applications Future Development Status as
Freeware The feedback driven development cycle
that produced gPRIME is ongoing (physically) at
the Howard Hughes Medical Institutes Janelia
Farm Research Campus in response to user needs
around the world. Future versions of the
software will have expanded hardware support and
more powerful analysis tools and will be hosted
by HHMI as freeware for Educational and
University Based Research Applications. A
licensing fee will be required for commercial
applications of the software.
Sorting by Peak Frequency of EOD in the Electric
Fish When recording electric organ discharges
from fish in a tank (or in a river), amplitude of
spikes is not a reliable measure given the
constant motion of the fish relative to the
electrode. The peak frequency component of the
discharge, however, is constant (unless
intentionally modified by the fish) and may be
used to visualize the location of distinct
individuals. Shown here is a courtship display
(provided by Carl Hopkins, Cornell University)
Intracellular Study of Synaptic Plasticity in the
Crayfish NMJ This study illustrates the ability
of gPRIME to measure Post Synaptic Potential
(PSP) amplitudes in an intracellular application.
Note how the amplitude of the first PSP is
smaller than the PSP at 6 seconds post tetanic
stimulation, thus illustrating a form of short
term potentiation.
6-Neuron Nerve Recordings (Extracellular) in the
Crayfish gPRIME allows for clustering of detected
events in a 2 parameter space. In this case,
Action Potentials where threshold detected and
sorted based on their signal energy density and
maximum amplitude.
Stretch Receptor (MRO) Adaptation in the
Crayfish Here, the exponential adaptation rate of
a stretch detecting proprioceptor neuron is
evident in response to a stretch and hold of the
muscle of interest. Students are able to
visualize this adaptation in real-time for an
immediate understanding of the underlying
physiological mechanisms.
Applications Around the World in Research and
Education While we have not actively tracked the
dissemination of this software to research groups
and educational environments, we know that it has
found application in labs in Spain, Venezuela,
and in several places in the United States
outside of Cornell University. One particular
application which we are actively involved in is
a High School collaboration between the Loudoun
County (Virginia) High School Academy of Science
(an HHMI Funded program) and a High School in
Singapore. Students will use gPRIME to turn
their laptop computers into data acquisition and
analysis tools (using the native sound card and a
microphone) to study the characteristics of
cricket chirps from a variety of species on both
sides of the globe. This free tool is currently
enabling this international collaboration and
connecting kids in a manner never before achieved.
Signal Conditioning Data Analysis in Real Time
and Offline gPRIME offers several tools for
signal conditioning and event detection. The
user may access built-in high and low pass
filters at a variety of frequencies between 10Hz
and 40kHz as well as a 60Hz notch filter for line
noise reduction. The user may also create their
own FIR filter coefficients and load them into
the program for application to the current signal
trace. Resulting events are initially detected
by amplitude thresholds and events may be
rejected based on a second amplitude threshold or
thresholds in energy density or peak frequency
component (of the FFT) space. The resulting
events are characterized in terms of Max Value,
Minimum Value, Event Time, Interval/Rate since
last event, window energy density, and signal
peak frequency component.
Publication Quality Graphics Production Any trace
may be extracted from any graphics window in the
program (see the small maroon box in the corner
of each axis containing graphics). The user has
the ability to overlay multiple traces in a
single figure, scale the figure and the data, and
control the properties of the resulting figure
(line width, color, order, data markers,
log/linear scale, etc with limited annotation
features). The resulting figures may be saved in
common graphics formats including TIFF and
JPEG. In addition to the graphics outputs,
gPRIME enables you to export results and raw data
in a variety of formats including .wav and ASCII
text files.