Skip to content
← Back to browse
BRAIN Publication-derived

Machine Learning-Enabled Detection of Electrophysiological Signatures in iPSC-Derived Models of Schizophrenia and Bipolar Disorder

Source Cheng et al., 2025 · Johns Hopkins University · 10.1063/5.0250559

👤 Kai Cheng, Autumn Williams, Anannya Kshirsagar, Sai Kulkarni, Rakesh Karmacharya, Deok-Ho Kim, Sridevi V. Sarma, Annie Kathuria ⏱ 270 days 📋 15 phases 🧫 Patient-Derived iPSC (Schizophrenia, Bipolar Disorder)

Abstract

This protocol describes the generation, maintenance, and electrophysiological characterization of patient-derived cerebral organoids (COs) and two-dimensional cortical interneuron cultures (2DNs) from individuals with schizophrenia and bipolar disorder. Multi-electrode array recordings combined with stimulus–response dynamic network modeling and machine learning enable identification of disease-specific electrophysiological signatures for objective psychiatric biomarker discovery.

Cell source
Patient-Derived iPSC (Schizophrenia, Bipolar Disorder)
Application
Disease modeling and biomarker identification

Protocol overview

53 steps across 15 phases

iPSC Line Preparation and Validation Passage 25-30
  1. 1 iPSC Genomic Integrity Testing with KaryoStat Analysis
  2. 2 Copy Number Variation Analysis
  3. 3 Mycoplasma Testing
Cerebral Organoid Differentiation Days 0–7 (EB formation and transition)
  1. 1 Culture iPSCs in Basal Medium
  2. 2 Form Embryoid Bodies in U-Bottom Plates
  3. 3 Transfer EBs to Cerebral Organoid Induction Media
  4. 4 Embed EBs in Matrigel
Cerebral Organoid Maturation Days 7–270 (9 months in vitro)
  1. 1 Culture Organoids on Orbital Shakers
  2. 2 Add BDNF Starting at Day 30
  3. 3 Continue Culture for 6–9 Months
Organoid Integrity Validation Multiple timepoints during culture (e.g., Days 30, 60, 90, 180, 270)
  1. 1 Immunohistochemical Characterization
  2. 2 Quantify Cell Populations
  3. 3 RT-qPCR Validation of Marker Expression
Two-Dimensional Cortical Neuron Differentiation Days 0–120 (neuronal maturation)
  1. 1 Culture iPSCs on Geltrex-Coated Plates
  2. 2 Neural Induction with N2/B27 and SMAD Inhibitors
  3. 3 Split and Plate NPCs on Day 8
  4. 4 Forebrain Specification with Purmorphamine
  5. 5 Plate on ECM-Coated Substrate
  6. 6 Neuronal Maturation in BrainPhys Media with DAPT
  7. 7 Continue Culture to Maturation
Preparation of Organoids for MEA Recording Day 270 (9 months in vitro, after MEA plate preparation)
  1. 1 Prepare MEA Plates with ECM Coating
  2. 2 Transfer Organoids to MEA Plates
  3. 3 Culture Organoids on MEA Plates for 3 Additional Months
Electrophysiological Recording Setup and Media Preparation Day of recording
  1. 1 Prepare Recording Media
  2. 2 Perform 50% Media Exchange 24 Hours Before Recording
  3. 3 Complete Media Exchange 1 Hour Before Recording
  4. 4 Set Up MEA Recording System with Temperature and pH Control
  5. 5 Verify Recording Media Osmolarity
Baseline and Post-Electrical Stimulation Recording Day of recording (9-month-old organoids or 90-120-day neurons)
  1. 1 Check Electrode Impedance and Integrity
  2. 2 Record Baseline Spontaneous Activity
  3. 3 Apply Electrical Stimulation Protocol
  4. 4 Record Post-Stimulation Response
Signal Validation and Spike Detection During and immediately after recording
  1. 1 Characterize Baseline Firing Properties
  2. 2 Apply Tetrodotoxin (TTX) to Verify Neuronal Origin
  3. 3 Verify Signal-to-Noise Ratio
Electrophysiological Data Preprocessing During data analysis (post-recording)
  1. 1 Apply Bandpass Filtering
  2. 2 Apply Notch Filtering at Power Line Harmonics
  3. 3 Spike Detection and Time Series Binarization
  4. 4 Downsample Spike Train to 1 kHz
  5. 5 Rate Code Spike Train with 200 ms Sliding Window
Stimulus–Response Dynamic Network Modeling (SRDNM) During data analysis
  1. 1 Develop SRDNM to Capture Stimulus-Response Relationships
  2. 2 Estimate State Transition Matrix A
Sink Index Feature Extraction During data analysis
  1. 1 Define Sink Index for Each Channel
  2. 2 Generate Comprehensive Sink Index Feature Map
Feature Selection Using Minimum Redundancy Maximum Relevance (MRMR) During data analysis
  1. 1 Apply MRMR Feature Selection Framework
  2. 2 Identify Top-Ranked Features for Classification
Support Vector Machine Classifier Development and Validation During data analysis
  1. 1 Implement Nested Cross-Validation Strategy
  2. 2 Train Support Vector Machine (SVM) Classifier
  3. 3 Evaluate SVM on Held-Out Test Fold
  4. 4 Aggregate Performance Metrics Across Outer Folds
Hyperparameter Tuning and Algorithm Comparison During data analysis
  1. 1 Optimize SVM Hyperparameters via Bayesian Optimization
  2. 2 Compare Performance of Multiple Classification Algorithms
  3. 3 Select SVM as Primary Classifier

Full SOP

🔬

Create a free account to access this protocol

Join OrganMatch to unlock step-by-step procedures, reagent concentrations, QC checklists, and downloadable batch record templates.

Create free account

Already registered? Log in

Attribution

This SOP was authored by Organthis based on the published method in Cheng et al., 2025. The originating laboratory holds no rights in this SOP and has not endorsed it unless marked Verified.

This wording is awaiting legal review.

Something wrong with this entry? Report an issue with this protocol

Need a commercial licence?
Use this protocol in your therapeutic or diagnostic pipeline.
Request Commercial Licence