Cell Types Data
Searching the Database
With the initial launch of the Allen Cell Types Database, we include electrophysiological recordings from 248 individual cells, a sub-set of which also include [morphological] reconstruction and [#Neuronal Models]. There are several methods to search this database; 1) enable specific [filter parameters] by selecting from the "Filters" menu, 2) Select a cell from the "[#Cell Location]" map, 3) use the slider bars in the [Parallel Coordinate Plot] or 4) selecting from one of the
curated search terms
only a sub-set of these cells are listed; however, you can alter the sub-set of cells represented by clicking on the "More Options+" link and selecting from categories that include "Dendrite Type", "Morphology", "Models" and "Apical Dendrite".
To filter the neurons in your list of results select from the [mouse line], the cortical layer (2/3, 4, 5, 6a), and the hemisphere (left, right or either).
You can also
the list of experimental results by the electrophysiological or morphological [features].
Whole cell current clamp recordings were made from cells expressing the fluorescent molecule, tdTomato. Fluorescent cells were cre-positive cells from one of the transgenic lines described below.
Rorb-IRES2-Cre-Strong expression in the zonal layer of the superior colliculus and subregions of thalamus. Dense, patchy expression in layer 4 and sparse expression in layer 5 and 6 in cortex. Also expressed in trigeminal nucleus and small patches of cells in cerebellum. View transgenic characterization.
Scnn1a-Tg2-Cre-Reporter expression in sparse and/or restricted regions of cortex (layer 4), thalamus, midbrain, medulla, pons, and cerebellum. View transgenic characterization.
Scnn1a-Tg3-Cre-Enriched in cortical layer 4 and in restricted populations within cortex, thalamus, and in cerebellum. View transgenic characterization.
Nr5a1-Cre-Expressed in restricted populations within the hypothalamus (ventromedial hypothalamus), and in cortical layer 4. View transgenic characterization.
Rbp4-Cre_KL100-Enriched in cortical layer 5 and the dentate gyrus. View transgenic characterization.
Ntsr1-Cre-Specific to cortical layer 6 neurons. View transgenic characterization.
Sst-IRES-Cre-Strong scattered expression throughout the brain. Localized areas of enrichment include restricted populations in thalamus, amygdala, midbrain, hindbrain, cortical subplate, and Purkinje cell layer. View transgenic characterization.
Pvalb-IRES-Cre-Expressed in restricted and/or sparse populations within the cerebellum, medulla, pons, midbrain, cortex, hippocampus, thalamus, and striatum. View transgenic characterization.
Htr3a-Cre_NO152-Reporter expression is detected in a subset of cortical interneurons. Enrichment is also detected in restricted populations within olfactory areas, pallidum, hypothalamus, pons, medulla, and cerebellum. View transgenic characterization.
Gad2-IRES-Cre-Specific to GABAergic neurons. Enriched in the striatum, piriform cortex, and within restricted populations in the thalamus, hypothalamus, cerebellum, olfactory areas, and GABAergic interneurons of the cortex. View transgenic characterization.
Cells were registered to the Allen Mouse Common Coordinate Framework (see whitepapers in "Documentation") and are illustrated in the approximate location on the representation of the visual cortical areas (see image below). Cell location in a two dimensional plane are represented in the center image view (from above) and the depth of the cell being recorded from is indicated in either the coronal or the saggital views which are shown in the bottom and side views, respectively. The coronal and saggital views are labeled with the Pia (top of the brain), the White Matter (or bottom of the cortex) as well as the Lateral (L), Medial (M), Rostral (R) or Caudal (C) perspectives of the brain.
Filtering the cells (using the [#Filters] parameters or the sliders in the [Parallel Coordinate Plot]) will result in a decrease or increase in the number of cells visualized on this map.
Once you have selected a cell (by clicking on a circle in the map) meta-data for that cell will be displayed in the upper right hand corner of the map.
Parallel Coordinate Plot of the Cell Features
To be able to visualize this data, which has many distinct and sometimes unrelated dimensions, we used a Parallel Coordinate Plot. This visualization plots each cell according to 6 electrophysiological or morphological features.
When you have selected the "
" Mode, six features are illustrated in the Parallel Coordinate Plot: Upstroke/Downstroke Ratio, Fast AP Trough, FI Curve Slope, Rheobase, Ramp AP Time and Resting Vm. When the "
Electrophysiology + Morphology
" Mode is selected, again only six features are shown: four electrophysiology features; Upstroke/Downstroke Ratio, FI Curve Slope, Rheobase and Resting Vm, as well as two Morphology features: Max Distance and # Stems.
You can limit your search by using the slider bars on each axis. Hovering your mouse over the axis of the parameter you are interested in inspecting will illuminate the data included in your search. Click and drag from the extremes of the axis to limit your search.
Each of the axes can be relocated by clicking and dragging the title to another position along the x-axis.
Cell Feature Filters
While many distinct electrophysiological features were recorded and/or calculated from each cell, only a few features were called out in the web application. To see the other features, please read the Electrophysiology whitepaper in the "Documentation" tab and/or download the data from the Allen SDK. These features were selected not only by what they were able to tell us about the intrinsic properties of the cells, but also in their ability to differentiate between cell-types in a non biased clustering analysis.
The ratio between the absolute values of the action potential peak upstroke and the action potential peak downstroke. Action potential peak upstroke: The maximum rate of change between the action potential threshold and the action potential peak. Action potential peak downstroke: The minimum rate of change between the action potential peak and the action potential trough.
This value gives us an indication of the amount of time a cell takes to recover after an Action Potential. An example of the difference between a high and a low ratio is demonstrated here.
This parameter is used to distinguish between spiny (putatively excitatory) and aspiny (putatively inhibitory) groups of cells in a non biased clustering analysis.
Fast AP Trough
Minimum value of the membrane potential in the interval lasting 5 ms after the peak of the initial action potential. Notice in the example above of a High vs. Low ratio, the Fast AP Trough values are related.
This value gives us an indication of the kinds of Potassium ion channels that are present in a cell and can be an indication of how fast the neuron can recover after an action potential. This value helps contribute to separating spiny and aspiny cells when used in combination with upstroke:downstroke.
FI Curve Slope
While some of the features we have represented in the application probe the firing patterns of single action potentials, we also calculated features based on multiple sweeps of the cell. The F/I Curve Slope is the slope calculated in the dynamic range of the frequency of action potential firing as a function of the current injected into the cell (in pA). This value indicates the excitability and the firing pattern of a cell.
This value was instrumental in distinguishing the aspiny cells (i.e. it clustered the Pvalb cells)
This value is the amount of stimulus current (long square current in pA) required to initiate a single action potential. This parameter gives us an indication of the excitability and the membrane resistance of the cell.
Ramp AP Time
This value is the time and voltage required to initiate a single action potential by a steady ramp of current injection (25 pA/sec). This value, while similar to Rheobase, was collected in a slightly different manner and in unbiased cluster analysis of cell-types was the best parameter to differentiate spiny neurons.
The resting membrane potential was measured soon after breaking into the cell and measures the membrane potential of a neuron rests with no applied current.
A sub-set of cells with electrophysiological recordings were manually selected to enter our automated 3-D reconstruction pipeline. Similar to the electrophysiology features, many morphology features were calculated in the reconstruction, but only two are called out in the web application; Max Distance and # Stems. To see the entire array of morphological features, please read the whitepaper in "Documentation" or download the data from the Allen SDK.
The two features represented in the web application were the most useful for discriminating between a particular category of spiny neuron ("tall, tufted") and a particular category of aspiny neuron ("common type"). These morphology categories are some of the most common in this dataset.
This is the maximum
of all the nodes. Euclidean distance is the straight line distance from the soma (root) to the node.
Number of Stems
number of stems
attached to the soma. This value does not include the apical dendrite.
Filtering of the dataset results in a list of experiments ranked by Upstroke:Downstroke. You can alter the sort parameters from the [#Filters] menu.
Different sets of stimulation waveforms were used in order to:
- Interrogate intrinsic membrane mechanisms that underlie the input/output function of neurons
- Linear and non-linear subthreshold properties
- Action potential initiation and propagation
- Understand aspects of neural response properties in vivo
- Stimulation frequency dependence (theta vs. gamma) of spike initiation mechanisms
- Ion channel states due to different resting potentials in vivo
- Construct and test computational models of varying complexity emulating the neural response to stereotyped stimuli
- Generalized leaky-integrate-and-fire (GLIF) models
- Biophysically and morphologically realistic conductance-based compartmental models
The Allen Cell Types Database contains two types of neuronal models: perisomatic biophysical models and generalized leaky integrate-and-fire (GLIF) models. These models attempt to mathematically reproduce a cell's recorded response to a current injection. The perisomatic biophysical models take into account dendritic morphological structure, whereas GLIF models are simple point neuron models which represent the neuron as a single compartment.
There are five levels of GLIF models with increasing levels of complexity. The most basic model is a simple leaky integrate-and-fire equation. More advanced GLIFs attempt to model variable spike threshold, afterspike currents, and threshold adaptation.
1. Leaky Integrate and Fire (LIF)
Standard circuit representation of a resistor and capacitor in parallel with a leaky membrane.
2. LIF + Reset Rules (LIF-R)
LIF with biologically-derived threshold and voltage reset rules in addition to a biologically derived threshold decay.
3. LIF + Afterspike Currents (LIF-ASC)
LIF with spike-induced currents to model long-term effects of voltage-activated ion channels.
4. LIF-R + Afterspike Currents (LIF-R-ASC)
LIF with additional Reset Rules and Afterspike Currents.
5. LIF-R-ASC + Threshold Adaptation (LIF-R-ASC-A)
All of the above, with an additional voltage-dependent component of threshold.
Models with active conductances at the soma and passive dendritic morphology based on full 3D reconstruction.