Sporadic ALS Initiation Biology: Deep Phenotyping of At-Risk Cohorts

PRIMARY OUTCOME

Identification of molecular signature predictive of ALS development within 2-3 years, combining genetic risk scores, biofluid biomarkers, and neurophysiological changes with >80% predictive accuracy.

EXPECTED OUTCOMES

# Expected Outcomes ## Primary Molecular Signatures **Novel Biomarker Panel**: Identification of 4-6 reproducible molecular signatures distinguishing pre-symptomatic ALS risk, potentially including: (1) elevated CSF phosphorylated tau species (pTau181/pTau217) with >2-fold enrichment in at-risk cohort; (2) neurofilament-based signature combining plasma NfL, pNfH, and phosphorylated tau ratios with longitudinal trajectory analysis; (3) inflammasome activation markers including CASP1 activity and cleaved gasdermin-D (GSDMD-NT) levels in blood and CSF; (4) circulating mitochondrial DNA burden with aberrant mtDNA/nuclear DNA ratio indicating senescent cell-derived mtDNA release; (5) metabolomic signature featuring dysregulated amino acid metabolism (reduced arginine, elevated kynurenine) and lipid peroxidation products; (6) neuroinflammatory cytokine profile with elevated IL-6, TNF-α, and MCP-1 demonstrating 1.5-3.0 fold enrichment compared to controls. ## Imaging and Neurophysiological Correlates **Structural and Functional Brain Changes**: Discovery of pre-symptomatic neuroimaging markers including reduced motor cortex grey matter volume (5-15% reduction), increased white matter tract diffusivity in corticospinal tracts (DTI FA reduction of 10-20%), and altered resting-state functional connectivity in motor networks. Identification of brainstem atrophy and metabolic hypometabolism on 18F-FDG PET correlating with molecular risk signatures. **Neurophysiological Abnormalities**: Documentation of subclinical motor system dysfunction including motor unit action potential (MUAP) complexity changes, reduced motor evoked potential amplitudes (10-30% reduction), increased intracortical facilitation ratios, and early denervation patterns on HD-sEMG preceding clinical weakness by 12-24 months. Correlate transcranial magnetic stimulation markers with biofluid neurofilament levels. ## Multi-Omics Integration Results **Transcriptomic Dysregulation**: Characterization of 50-100 dysregulated genes in motor cortex including downregulation of synaptic plasticity genes, upregulation of stress response pathways (HSPA, HSP90, UPR components), and altered expression of RNA processing factors (TARDBP targets, FUS-interacting proteins). Mapping of alternative splicing anomalies in disease-relevant transcripts. **Metabolic Reprogramming**: Evidence of motor neuron-selective metabolic dysfunction including reduced oxidative phosphorylation efficiency, increased glycolytic dependence, dysregulated amino acid catabolism, and accumulation of lipid peroxidation products. Identification of specific metabolites (e.g., kynurenine, glutamate, branched-chain amino acids) with predictive utility. **Inflammasome and Innate Immune Activation**: Demonstration of low-grade, compartmentalized inflammasome priming in peripheral immune cells and elevated CASP1 activity associated with pyroptotic cell death markers. Evidence of microbial lipopolysaccharide (LPS) translocation and pattern recognition receptor (TLR4, TLR2) hyperresponsiveness in at-risk individuals. ## Predictive Machine Learning Model **Algorithm Performance**: Development of integrated predictive model achieving ≥80% accuracy, ≥75% sensitivity, ≥85% specificity for ALS development within 2-3 years in held-out test cohort. Generation of individualized risk scores stratifying participants into low-risk (<5% 3-year ALS probability), intermediate-risk (5-25%), and high-risk (>25%) categories. Demonstration of superior discriminative performance (AUC 0.88-0.95) compared to current clinical risk assessment methods (AUC 0.65-0.75). ## Mechanistic Insights **Pathway Integration**: Mapping of integrated stress response pathways showing convergence between senescent cell mitochondrial DNA release, inflammasome activation, axonal RNA transport dysfunction, and R-loop accumulation. Identification of potential intervention points and biomarker-guided therapeutic targets. Provisional ranking of hypothesis validation with quantitative support from multi-omics data.

SUCCESS CRITERIA

# Success Criteria ## Molecular Signature Validation **Statistical Rigor**: Achievement of p<0.05 (after Bonferroni correction for multiple comparisons) in differential expression/abundance analysis comparing pre-symptomatic at-risk individuals to age-matched controls for each proposed molecular signature. Demonstration of effect sizes (Cohen's d or log2 fold-change) ≥0.8 for primary biomarkers. Confirmation of findings using independent statistical approaches (parametric t-tests, non-parametric Mann-Whitney U tests, linear mixed-effects models accounting for repeated measures). **Reproducibility Across Cohorts**: External validation demonstrating ≥60% concordance of identified molecular signatures in independent validation cohorts (n≥100) recruited from different geographic regions/institutions. Successful replication of top 5 identified biomarkers with consistent direction and magnitude of effect across ≥2 independent datasets. Publication of results in peer-reviewed journals with openly shared code and standardized protocols. **Temporal Dynamics**: Documentation of longitudinal biomarker trajectories showing progressive changes in molecular signatures during pre-symptomatic period, with quarterly or semi-annual measurement intervals. Identification of individuals demonstrating rapid biomarker progression with subsequent clinical ALS diagnosis, validating predictive utility. Demonstration that baseline biomarker levels significantly predict rate of change over follow-up period (p<0.05). ## Predictive Model Performance **Primary Accuracy Benchmark**: Attainment of ≥80% overall accuracy on held-out test set (n≥84 at-risk + 63 controls) using primary integrated machine learning model. Maintenance of ≥75% sensitivity (true positive rate) to minimize false negatives among truly at-risk individuals. Maintenance of ≥85% specificity (true negative rate) to minimize unnecessary clinical interventions. Documentation of positive predictive value (PPV) ≥70% and negative predictive value (NPV) ≥90% for clinical utility. **Secondary Performance Metrics**: Generation of area under receiver operating characteristic curve (AUC-ROC) ≥0.88 on test cohort. Achievement of area under precision-recall curve (AUC-PR) ≥0.80 reflecting performance in imbalanced classification scenario. Demonstration of model calibration with calibration slope and intercept within acceptable ranges (slope 0.85-1.15, intercept -0.10 to +0.10). **Stratification and Risk Stratification**: Successful stratification of cohort into risk tiers with statistically significant differences in ALS development rates between groups (log-rank test p<0.001 for Kaplan-Meier curves). Achievement of ≥5-fold relative risk between high-risk and low-risk groups. Generation of individual probability estimates with narrow 95% confidence intervals reflecting model certainty. ## Clinical and Mechanistic Validation **Outcome Prediction**: Prospective validation that ≥80% of individuals assigned to high-risk category (top quartile of model probability scores) develop clinical ALS within 2-3 year follow-up window. Conversely, demonstration that <10% of low-risk category individuals develop ALS during same timeframe. Correlation of model predictions with objective clinical progression measures (rate of ALSFRS-R decline, neurophysiological deterioration rate). **Mechanistic Coherence**: Demonstration that identified molecular signatures mechanistically cohere around proposed biological pathways (senescent cell mitochondrial DNA release, inflammasome activation, axonal transport dysfunction) through pathway enrichment analysis (p<0.001) and functional validation studies. Evidence from iPSC-derived motor neuron models showing recapitulation of identified molecular phenotypes. Documentation of pathway biomarker interactions with quantified epistatic effects. **Superiority to Current Methods**: Statistical proof (DeLong's test p<0.05) that integrated multi-omics model demonstrates significantly superior predictive performance compared to: (1) single biomarker approaches, (2) genetic risk scores alone, (3) clinical risk assessment scales, and (4) conventional neuroimaging markers. Generation of clinical decision curves demonstrating net benefit of new model across relevant risk thresholds. ## Data Quality and Standardization **Assay Reproducibility**: Achievement of intra-assay coefficient of variation (CV) <10% and inter-assay CV <15% for all quantitative biomarker measurements. Demonstration of platform concordance for overlapping measurements (e.g., >0.95 Pearson correlation between different proteomics platforms). Implementation of blinded sample analysis and randomized batch processing to minimize systematic bias. **Analytical Standards**: Adherence to STROBE guidelines for observational studies, REMARK guidelines for biomarker studies, and transparent reporting of machine learning development (TRIPOD checklist). Pre-specification of statistical analysis plan prior to outcome assessment. Documentation of all data preprocessing steps, transformation procedures, and quality control filters with predetermined thresholds.

PROTOCOL

# Sporadic ALS Initiation Biology: Deep Phenotyping Protocol ## Phase 1: Cohort Recruitment and Baseline Characterization (Months 1-12) **Participant Selection and Stratification**: Recruit 500 at-risk individuals (age 40-75) meeting criteria: (1) first-degree relatives of sALS patients (n=200), (2) occupational/environmental exposure history (n=150), (3) subclinical neurophysiological abnormalities on screening electromyography (n=100), (4) elevated serum phosphorylated tau or neurofilament light chain (>75th percentile; n=50). Enroll 300 age-matched, exposure-matched controls. Perform comprehensive clinical assessments including ALS Functional Rating Scale-Revised (ALSFRS-R), Medical Research Council grading, quantitative grip strength dynamometry, and detailed occupational/environmental exposure questionnaires. ## Phase 2: Multi-Modal Deep Phenotyping (Months 3-24) **Neuroimaging Protocol**: Conduct 3T MRI brain and spinal cord imaging with quantitative T2 relaxometry, diffusion tensor imaging (DTI), and resting-state functional MRI at baseline, 12, and 24 months. Perform 18F-FDG PET imaging to assess motor cortex and brainstem metabolic activity. Measure cortical thickness, grey matter volume, and white matter tract integrity using automated segmentation pipelines (FSL, SPM12). **Advanced Neurophysiology**: Perform high-density surface electromyography (HD-sEMG) with motor unit action potential decomposition across upper and lower limbs. Conduct transcranial magnetic stimulation (TMS) with assessment of motor evoked potentials, silent period duration, and intracortical inhibition/facilitation ratios. Measure sensorimotor conduction velocity and F-wave chronodispersion. Repeat at 6-month intervals. **Biofluid Collection and Analysis**: Obtain venous blood (40 mL), cerebrospinal fluid via lumbar puncture (12 mL), and saliva at baseline and every 6 months. Quantify phosphorylated tau-181 (pTau181), phosphorylated tau-217 (pTau217), neurofilament light chain (NfL), phosphorylated neurofilament heavy chain (pNfH), and glial fibrillary acidic protein (GFAP) using ultrasensitive immunoassays (Simoa HD-X platform). Measure inflammatory cytokines (IL-6, TNF-α, IL-1β, IL-10, MCP-1) using high-sensitivity multiplex assays. Assess extracellular vesicle concentration and phosphorylated protein content via nanoparticle tracking analysis. **Metabolomic and Lipidomic Profiling**: Perform untargeted liquid chromatography-mass spectrometry (LC-MS/MS) metabolomics on plasma and CSF, quantifying >1,200 metabolites across amino acid, lipid, carbohydrate, and nucleotide pathways. Conduct high-resolution lipidomics measuring >600 lipid species using LC-MS/MS with class-specific fragmentation. Apply stable isotope tracer studies (13C-glucose, 15N-amino acids) to assess metabolic flux through glycolysis, TCA cycle, and amino acid catabolism. **Immunological Assessment**: Perform comprehensive immunophenotyping via 18-color flow cytometry quantifying T cell subsets (CD4+, CD8+, Treg, Th17), B cell populations, monocyte subsets, and NK cells. Measure ex vivo cytokine production following lipopolysaccharide (LPS) stimulation. Assess mitochondrial DNA (mtDNA) release using ddPCR quantification of circulating mtDNA copies. Evaluate inflammasome activation markers (ASC specks, pro-caspase-1) in peripheral blood mononuclear cells (PBMCs) using immunofluorescence and flow cytometry. ## Phase 3: High-Throughput Multi-Omics Integration (Months 12-36) **Genomic and Transcriptomic Analysis**: Perform whole exome sequencing (WES) at 100x coverage with variant annotation for known ALS genes (SOD1, FUS, TARDBP, C9orf72, VCP, UBIQUILIN2, OPTN, ATXN2) and emerging candidates. Conduct targeted RNA-sequencing on motor cortex biopsies (n=150 participants) using 3' digital counting methodology, quantifying expression of 500 ALS-relevant genes. Assess alternative splicing patterns in TARDBP and FUS targets. Measure C9orf72 repeat expansion status via repeat-primed PCR and digital droplet PCR. **Proteomic Profiling**: Perform data-independent acquisition (DIA) mass spectrometry on CSF and plasma, quantifying >2,000 proteins across 4 sample timepoints. Conduct targeted proteomics on CASP1, GSDMD, pro-IL-1β, and related inflammasome components. Measure protein phosphorylation sites (phosphoproteomics) in PBMC lysates using tandem mass tag (TMT) quantification. **Functional Validation**: Derive induced pluripotent stem cells (iPSCs) from 50 at-risk individuals and 25 controls. Differentiate to motor neurons using dual SMAD inhibition protocol. Assess mitochondrial membrane potential, calcium handling, and response to oxidative stress. Measure CASP1 activity, inflammasome complex formation, and pyroptotic cell death markers (LDH release, propidium iodide uptake). ## Phase 4: Machine Learning Model Development (Months 24-36) **Data Integration and Feature Selection**: Integrate 50+ molecular, imaging, and neurophysiological variables into unified analytical framework. Apply recursive feature elimination and LASSO regression for dimensionality reduction. Normalize data using quantile normalization and ComBat batch correction. **Algorithm Development**: Develop ensemble machine learning models (random forest, gradient boosting, neural networks) trained on 70% of cohort (n=280 at-risk + 210 controls) with 30% held-out test set. Implement 5-fold cross-validation with stratification by recruitment stratum. Generate receiver operating characteristic (ROC) curves and calculate area under curve (AUC). Optimize for sensitivity/specificity trade-off targeting ≥80% accuracy, ≥75% sensitivity, ≥85% specificity.

LINKED HYPOTHESES

Scoring Dimensions

Prerequisite Graph (4 upstream, 5 downstream)

Missions

HTML iframe

<iframe src="http://scidex.ai/artifact/exp-wiki-experiments-als-sporadic-initiation-biology?embed=1" width="100%" height="600" style="border:0;border-radius:8px"></iframe>

Markdown link

[Sporadic ALS Initiation Biology: Deep Phenotyping of At-Risk Cohorts](http://scidex.ai/artifact/exp-wiki-experiments-als-sporadic-initiation-biology)

Direct URL

http://scidex.ai/artifact/exp-wiki-experiments-als-sporadic-initiation-biology

Preview