Data

Downloads Page

This data and code are freely available to download.

If you make use of the data or the code, please reference the appropriate publications. Please also state clearly where and how you got the data or the code.

The MultiPatch Data Acquisition Software

The MultiPatch software — which runs in Igor Pro — is custom-made for carrying out spike-timing-dependent plasticity experiments with quadruple whole-cell recordings. Its development was begun in Sept 1999. Essentially all primary research papers by Dr Sjöström relied on MultiPatch. This software is currently provided as-is and without a manual.

The qMorph Morphometry Analysis Software

In association with the Zhou et al Scientific Reports 2021 11(1):12695 paper, we provide the qMorph code for analysis of morphological reconstructions of neurons. We previously used the qMorph morphometry code for e.g. the Buchanan et al Neuron 2012 and the Lalanne et al JPhysiol 2016 studies.

Online Neuroscience Teaching Tools

During the summer of 2020, the undergraduate summer intern Chengxin Yu made code running online to help with understanding key neuroscience concepts taught in the course PHGY 311 - Channels, Synapses, & Hormones. These include the Hodgkin-Huxley model of action potentials, the leaky integrate-and-fire neuron model, a vesicle depletion model of short-term plasticity, and a binomial model of synaptic release.

CV Analysis Tutorial Software

In Brock et al Frontiers in Synaptic Neuroscience 2020 12:11, we provide code for simulating pitfalls associated with using CV analysis to assess the locus of expression of long-term plasticity.

Synaptic Quantal Parameters

In Costa et al Neuron 2017 96(1):177-189.e7, quantal parameters p and q were extracted from Sjöström et al, Neuron 2001 as well as from Sjöström et al, Neuropharmacology 2007. The estimated quantal parameters for individual plasticity experiments are available at Mendeley.

Quantal Parameters in Synaptic Plasticity

In Costa et al eLife 2015;4:e09457, average quantal parameters were extracted in STDP. Estimated changes in p, q and mean weight are available to download at Dryad. The biologically tuned computer model is available at ModelDB.

Dendritic Location Dependence of Plasticity

Location Dependence of Synaptic Plasticity: See Fig 3 of Sjöström and Häusser, Neuron 2006. Baseline period was ten minutes, followed by the 2.3-min-long induction period. The same induction protocol was used in all cases: five spikes at 50 Hz in pre- and postsynaptic neurons, temporally shifted at +10 ms, repeated 15 times. Inset graph shows the same results but with respect to the Euclidian distance between the putative synapse and the soma. Note that the dendritic path length was not measured, nor the dendrite branch thickness.
Formats: Excel | Igor saved graph 1 | Igor saved graph 2 | PDF

Connectivity of Layer-5 Pyramidal Cells

Connectivity dataset

With plasticity experiments using quadruple recordings, many cells are not connected. Many connections can also not be used for plasticity experiments, simply because the connections are very weak, or maybe they simply go bad. Fortunately, each quad recording provides a lot of data on the local connectivity. In the Song et al, PLoS Biology 2005 study, we made use of a relatively large quad-recording data set generated this way, consisting of more than 900 connections and 8000 cells. The data is arranged as described below. Format: Excel

Date: The date the experiment was conducted.

Attempt#: Each experimental day, I would attempt at getting a quad recording, and I would number each quad attempt.

nConnections: For each quad attempted, the nConnections parameter tells you how many connections I found.

nTested: For each quad attempt, I wouldn't necessarily get all four cells. If I got four, then I tested 12 connections, but if I only got three cells, then I tested 6 possible connections. With two cells, I tested two possible connections.

Age: The age of the animal, probably with a +/- one day slop or less.

CalciumConc: The concentration of calcium in the external solution. The reason this changes between 2.5 mM and 2.0 mM is because I modelled some of my studies on Feldman Neuron 2001 and Markram Science 1997. The former employed 2.5 mM, whereas the latter used 2.0 mM external calcium. The calcium concentration may affect both the amplitude and the CV of synaptic connections.

ConnectionString: Each quad recording has a ConnectionString associated with it, which uniquely identifies the connectivity pattern within the quad. For example, on 20/01/2001, Attempt #1, the string is 3_4,0.00020913,9.5405e-05; 4_3,0.00034122,0.00012075; 3_2,0.00050473,9.8416e-05; 2_3,0.00060293,0.00011659; 2_4,0.00060349,0.00013282; 1_4,0.0005404,0.00010957;. This means cell three was connected to cell 4 with a connective strength of 0.20913 mV and the standard deviation of that connection was 9.5404E-5 V. That pair also had a reciprocal connection from cell four to cell three, of 0.34122 mV strength, and so on and so forth. If the string says NA, then there was no connection found.

Note that all bad recordings were discared before this analysis stage, so recorded cells that were deemed bad were not included in the table -- this is important for obtaining correct connectivity counts.