Discovering Effective Connectivity in Neural Circuits: Analysis Based on Machine Learning Methodology

Abstract

As multielectrode array technology increases in popularity, accessible analytical tools become necessary. Simultaneous recordings from multiple neurons may produce huge amounts of information. Traditional tools based on classical statistics are either insufficient to analyze multiple spike trains or sophisticated and expensive in computing terms. In this communication, we put to the test the idea that AI algorithms may be useful to gather information about the effective connectivity of neurons in local nuclei at a relatively low computing cost. To this end, we decided to explore the capacity of the algorithm C5.0 to retrieve information from a large series of spike trains obtained from a simulated neuronal circuit with a known structure. Combinatory, iterative and recursive processes using C5.0 were built to examine possibilities of increasing the performance of a direct application of the algorithm. Furthermore, we tested the applicability of these processes to a reduced dataset obtained from original biological recordings with unknown connectivity. This was obtained in house from a mouse in vitro preparation of the spinal cord. Results show that this algorithm can retrieve neurons monosynaptically connected to the target in simulated datasets within a single run. Iterative and recursive processes can identify monosynaptic neurons and disynaptic neurons under favorable conditions. Application of these processes to the biological dataset gives clues to identify neurons monosynaptically connected to the target. We conclude that the work presented provides substantial proof of concept for the potential use of AI algorithms to the study of effective connectivity.

Synchronous firing of dorsal horn neurons at the origin of dorsal root reflexes in naïve and paw-inflamed mice

“…The outcome of this analysis reinforces the previous view that some presynaptic neurons triggered action potentials in the afferents and these, in turn, triggered the firing of post-synaptic neurons. The algorithm has been tested in complex synthetic circuits with good results (Pozo-Jimenez et al, 2021) and it is likely to detect close functional connectivity such as that produced by monosynaptic connections. Putative presynaptic neurons detected were of the IFB and IS types, which again, suggest a particularly relevant role for these neurons in the generation of backfiring and population bursts.…”

“…Briefly, after deep anesthesia (urethane, 2 g/Kg i.p. ), a lumbar laminectomy was performed and the cord dissected and placed in sucrose-substituted artificial cerebrospinal fluid (ACSF) at 4 • C. Then, meninges were removed and, in some spinal cords, L4 dorsal root was gently teased to obtain thin rootlets to record individual primary afferents, as previously reported (Pozo-Jimenez et al, 2021). Finally, we obtained a single horizontal slice of about 500 µm thick using a vibratome (Sectioning Systems 1500), which contained the dorsal horn and dorsal roots attached.…”

“…For the analysis of effective connectivity between neurons and primary afferents, we used a machine learning-based tool constructed with the algorithm C5.0 (Quinlan, 1993) that we developed, explained in depth and tested in synthetic and biological circuits in a previously published work (Pozo-Jimenez et al, 2021).…”

“…This score is calculated from the results of the confusion matrix obtained when the predictions made by the algorithm are compared with actual data during the testing phase. The maximum value for the Matthew's correlation coefficient (MCC) is 1 and values ≥ 0.18 for single neurons as attributes have been associated to monosynaptic excitatory connections using complex synthetic circuits in which more than one neuron is required to trigger action potentials in the target (Pozo-Jimenez et al, 2021). Therefore, here we used this value of MCC as indicative of a strong functional relation between neurons.…”