Raster plots machine learning to predict the seizure liability of drugs and to identify drugs

Matsuda, Naoki; Odawara, Aoi; Kinoshita, Koshi; Okamura, Ai; Shirakawa, Takafumi; Suzuki, I.

doi:10.1038/s41598-022-05697-8

Cited by 14 publications

(10 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In many studies, ASD efficacy is inferred through changes in certain features relevant to ictogenesis, such as burst firing or measures of network activity propagation. A recent study demonstrated that a neural network could be trained on raster plots of spiking activity to distinguish between antiseizure and seizure‐provoking compounds 45 . Although the impact of different ASDs upon individual network features has been examined, 18,39 our study is unique by virtue of the wide range of features and the use of dimensionality reduction techniques, and because our primary aim is ASD classification rather than proof of therapeutic efficacy.…”

Section: Discussionmentioning

confidence: 99%

“…A recent study demonstrated that a neural network could be trained on raster plots of spiking activity to distinguish between antiseizure and seizure-provoking compounds. 45 Although the impact of different ASDs upon individual network features has been examined, 18,39 our study is unique by virtue of the wide range of features and the use of dimensionality reduction techniques, and because our primary aim is ASD classification rather than proof of therapeutic efficacy. As such, we have not drawn conclusions in this work regarding the therapeutic application of drugs based on changes in network activity.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Classification of antiseizure drugs in cultured neuronal networks using multielectrode arrays and unsupervised learning

Bryson

Mendis²,

Morrisroe

et al. 2022

Epilepsia

View full text Add to dashboard Cite

Objective: Antiseizure drugs (ASDs) modulate synaptic and ion channel function to prevent abnormal hypersynchronous or excitatory activity arising in neuronal networks, but the relationship between ASDs with respect to their impact on network activity is poorly defined. In this study, we first investigated whether different ASD classes exert differential impact upon network activity, and we then sought to classify ASDs according to their impact on network activity. Methods:We used multielectrode arrays (MEAs) to record the network activity of cultured cortical neurons after applying ASDs from two classes: sodium channel blockers (SCBs) and γaminobutyric acid type A receptor-positive allosteric modulators (GABA PAMs). A two-dimensional representation of changes in network features was then derived, and the ability of this low-dimensional representation to classify ASDs with different molecular targets was assessed.Results: A two-dimensional representation of network features revealed a separation between the SCB and GABA PAM drug classes, and could classify several test compounds known to act through these molecular targets. Interestingly, several ASDs with novel targets, such as cannabidiol and retigabine, had closer similarity to the SCB class with respect to their impact upon network activity.Significance: These results demonstrate that the molecular target of two common classes of ASDs is reflected through characteristic changes in network activity of cultured neurons. Furthermore, a low-dimensional representation of network features can be used to infer an ASDs molecular target. This approach may allow for drug screening to be performed based on features extracted from MEA recordings.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Classification of antiseizure drugs in cultured neuronal networks using multielectrode arrays and unsupervised learning

Bryson

Mendis²,

Morrisroe

et al. 2022

Epilepsia

View full text Add to dashboard Cite

show abstract

“…Several in vitro and in vivo analysis platforms with medium throughput have been used for the early detection of the seizure liability (ictogenicity) of novel drug candidates [ 4 ]. These platforms include brain slice and multi-electrode array analysis of drug-induced epileptiform activity in rodent cortical neurons or slices, or human induced pluripotent cells [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 ], and the analysis of motor behavior generated by zebrafish larvae exposed to convulsant drugs [ 15 , 16 ]. Despite the promising data and increasingly stronger predictive values achieved by applying machine-learning analysis approaches, the predictive accuracy remains limited or has not yet been vigorously validated in different laboratories [ 17 ].…”

Section: Introductionmentioning

confidence: 99%

Gene Expression Profile as a Predictor of Seizure Liability

Lipponen

Kajevu

Natunen

et al. 2023

IJMS

View full text Add to dashboard Cite

Analysis platforms to predict drug-induced seizure liability at an early phase of drug development would improve safety and reduce attrition and the high cost of drug development. We hypothesized that a drug-induced in vitro transcriptomics signature predicts its ictogenicity. We exposed rat cortical neuronal cultures to non-toxic concentrations of 34 compounds for 24 h; 11 were known to be ictogenic (tool compounds), 13 were associated with a high number of seizure-related adverse event reports in the clinical FDA Adverse Event Reporting System (FAERS) database and systematic literature search (FAERS-positive compounds), and 10 were known to be non-ictogenic (FAERS-negative compounds). The drug-induced gene expression profile was assessed from RNA-sequencing data. Transcriptomics profiles induced by the tool, FAERS-positive and FAERS-negative compounds, were compared using bioinformatics and machine learning. Of the 13 FAERS-positive compounds, 11 induced significant differential gene expression; 10 of the 11 showed an overall high similarity to the profile of at least one tool compound, correctly predicting the ictogenicity. Alikeness-% based on the number of the same differentially expressed genes correctly categorized 85%, the Gene Set Enrichment Analysis score correctly categorized 73%, and the machine-learning approach correctly categorized 91% of the FAERS-positive compounds with reported seizure liability currently in clinical use. Our data suggest that the drug-induced gene expression profile could be used as a predictive biomarker for seizure liability.

show abstract

“…Thus, a successful data acquisition in MEA assays might be costly and time-consuming. Finally, an in vitro to in vivo extrapolation (IVIVE) is required in MEA using hiPSC-derived neurons [28][29][30][31]. Traditional MEA measurements are preferred to analyze spikes with frequency components of 1 kHz or more and to investigate pathological mechanisms or drug efficacy [32].…”

Section: Introductionmentioning

confidence: 99%

A functional neuron maturation device provides convenient application on microelectrode array for neural network measurement

et al. 2022

View full text Add to dashboard Cite

Background Microelectrode array (MEA) systems are valuable for in vitro assessment of neurotoxicity and drug efficiency. However, several difficulties such as protracted functional maturation and high experimental costs hinder the use of MEA analysis requiring human induced pluripotent stem cells (hiPSCs). Neural network functional parameters are also needed for in vitro to in vivo extrapolation. Methods In the present study, we produced a cost effective nanofiber culture platform, the SCAD device, for long-term culture of hiPSC-derived neurons and primary peripheral neurons. The notable advantage of SCAD device is convenient application on multiple MEA systems for neuron functional analysis. Results We showed that the SCAD device could promote functional maturation of cultured hiPSC-derived neurons, and neurons responded appropriately to convulsant agents. Furthermore, we successfully analyzed parameters for in vitro to in vivo extrapolation, i.e., low-frequency components and synaptic propagation velocity of the signal, potentially reflecting neural network functions from neurons cultured on SCAD device. Finally, we measured the axonal conduction velocity of peripheral neurons. Conclusions: Neurons cultured on SCAD devices might constitute a reliable in vitro platform to investigate neuron functions, drug efficacy and toxicity, and neuropathological mechanisms by MEA.

show abstract

Raster plots machine learning to predict the seizure liability of drugs and to identify drugs

Cited by 14 publications

References 46 publications

Classification of antiseizure drugs in cultured neuronal networks using multielectrode arrays and unsupervised learning

Classification of antiseizure drugs in cultured neuronal networks using multielectrode arrays and unsupervised learning

Gene Expression Profile as a Predictor of Seizure Liability

A functional neuron maturation device provides convenient application on microelectrode array for neural network measurement

Contact Info

Product

Resources

About