SYNAPTIK v0.9.0

SYNAPTIK v0.9.0: 3D RNA Biophysics (Turner 2004 Zuker MFE + steric competition), Patient Natural History Digital Twin (1,000-twin PRO-ACT/Answer ALS empirical Bayes Monte Carlo), Clinical Allometry Safety Guardrails & Protocol Corrections, Sugimoto 1995 Nearest-Neighbor Thermodynamic Verification, Literature-Anchored Off-Target Screening, and Concurrent Rate-Limited Literature Ingest. Author: Francesco Nobile (ORCID: 0009-0001-9477-3332).

SYNAPTIK v0.8.0 (Alpha): Patient VCF Ingestion, Drug Repurposing & FDA Pre-IND Package

SYNAPTIK v0.8.0 (Alpha): Patient VCF Ingestion, Drug Repurposing, SVG Visualizer & FDA Pre-IND Package

SYNAPTIK v0.8.0 brings four major clinical and regulatory capabilities to the translational neurogenetics platform, bridging raw patient sequencing directly to FDA-ready IND briefing documents.

Highlights & Key Additions

• Direct Patient Genomic VCF Ingestion & Prioritization Engine (synaptik vcf):
  • Ingests standard Variant Call Format (VCF v4.2 / v4.3) files (.vcf and .vcf.gz) directly from whole-genome (WGS) or whole-exome (WES) sequencing.
  • Automatically resolves ClinVar significance, gnomAD allele frequencies, CADD Phred scores, and SpliceAI donor/acceptor delta scores.
  • Prioritizes cryptic splice-altering variants across core neuro targets (STMN2, UNC13A, SMN2, ELP1, COL6A1, ATXN2, TARDBP, SOD1, FUS).
  • Direct pipeline bridge: synaptik nof1 --vcf <FILE.vcf> auto-extracts high-priority variants and runs the N-of-1 pipeline with zero manual formatting.
• Small-Molecule Drug Repurposing & CNS Tractability Engine (synaptik repurpose):
  • Curated database of brain-penetrant small-molecule drugs (Risdiplam, Branaplam, Riluzole, Edaravone, Sodium Phenylbutyrate, Memantine, Diazepam, Vorinostat, Rapamycin).
  • Evaluates Physicochemical CNS Multiparameter Optimization (CNS MPO, 0–6 scale) and Blood-Brain Barrier (BBB) penetrance.
  • Directional mechanism-of-action matching (Activator vs Inhibitor vs Modulator).
  • Predicts in vitro and in vivo transcriptomic rescue efficacy percentages and ranks candidates by composite translational rank score.
• Publication-Grade Splicing & ASO Architecture Vector Visualizer (SVG):
  • Generates publication-ready SVG diagrams rendering pre-mRNA architecture: canonical flanking exons, cryptic exon, splice donor (GT) and acceptor (AG) sites, sashimi-style curved splice junction arcs with $\Delta\Psi$, and tiled 18-mer ASO steric-blocker footprints with base-level sequences.
  • Standalone SVG file export via CLI flags: synaptik aso --svg <PATH.svg> and synaptik nof1 --svg <PATH.svg>.
  • Embedded directly into the laboratory Web UI for real-time visualization and one-click figure export.
• FDA CDER/CBER Type B Pre-IND Meeting Package Generator:
  • Expands synaptik nof1 with --pre-ind <PATH.md>: produces an audit-ready, 21 CFR Part 312 compliant 10-section eCTD briefing document.
  • Includes CMC specifications, nonclinical pharmacology, 13-week NHP GLP tox study plan, and specific questions for FDA reviewers.
• Interactive Laboratory Web Dashboard Expansion (synaptik ui):
  • Dedicated Patient VCF tab with text/file upload, sample STMN2 loader, and 1-click "Translate in N-of-1" routing.
  • Dedicated Drug Repurposing tab with interactive drug cards, CNS MPO scores, and clinical phase filters.
  • Embedded interactive vector SVG splicing visualizer with one-click SVG export.
  • One-click Download FDA Pre-IND Package button.
  • REST API endpoints: /api/vcf, /api/repurpose, /api/visualize/splicing, /api/nof1/preind.
• Scientific Recognition & Attribution:
  • Platform Lead: Francesco Nobile (ORCID: 0009-0001-9477-3332), Center for Genomic Medicine, Massachusetts General Hospital / Harvard Medical School & University of Pavia.
  • Standardized citation output via synaptik cite (BibTeX, APA, Nature format).
• CI Test Suite: 478 passing checks across 13 test suites on Python 3.10, 3.11, and 3.12.

SYNAPTIK v0.7.0 (Alpha): Autonomous N-of-1 Rare Disease & Translational Pipeline

SYNAPTIK v0.7.0 (Alpha): Autonomous N-of-1 Rare Disease & Translational Pipeline

SYNAPTIK v0.7.0 unifies patient genomic discovery and regulatory-ready therapeutic translation into an autonomous closed-loop platform.

Highlights

• Autonomous N-of-1 Rare Disease Pipeline (\synaptik nof1):
  • Splicing mechanism resolution for cryptic exon targets (STMN2, UNC13A, SMN2, ELP1, COL6A1, ATXN2, TARDBP).
  • In vivo human post-mortem single-cell concordance against human motor cortex snRNA-seq atlases.
  • In silico therapeutic ASO engineering with thermodynamic and specificity validation.
  • Causal genomic counterfactual simulation predicting network transcriptomic recovery.
  • Longitudinal fluid biomarker trajectory modeling (Plasma NfL) and patient clinical stratification.
  • Preclinical study design protocol generation with SHA-256 pre-registration fingerprint and 96-well randomized complete block layout.
  • Production of audit-ready, 7-page Markdown IND-enabling dossiers (--dossier ).
• Interactive Laboratory Web Dashboard & REST API Suite (\synaptik ui):
  • Web interface with dedicated N-of-1 Pipeline execution and citation export tabs.
  • REST API endpoints: /api/nof1, /api/aso, /api/translate, /api/simulate, /api/biomarker, /api/protocol, /api/cite.
• Author Recognition & Citation Framework (\synaptik cite):
  • Lead Author: Francesco Nobile (ORCID: 0009-0001-9477-3332), Center for Genomic Medicine, Massachusetts General Hospital / Harvard Medical School & University of Pavia.
• CI Test Suite: 406 passing checks across 10 test suites on Python 3.10, 3.11, and 3.12.

v0.6.1 (Alpha): Scientific Recognition & Citation Framework

SYNAPTIK v0.6.1: Scientific Recognition, Citation Engine, and Author Metadata

This release establishes the scientific citation framework and author attribution across SYNAPTIK, ensuring that whenever researchers, drug discovery teams, or academic labs use this platform, lead author Francesco Nobile and the tool SYNAPTIK are cited with full academic credit.

Highlights

• synaptik cite CLI Command & API:
  • Direct metadata querying and academic citation generation for lead author Francesco Nobile (ORCID: 0009-0001-9477-3332), Center for Genomic Medicine, Massachusetts General Hospital / Harvard Medical School & University of Pavia.
  • Native multi-format export:
    • synaptik cite --bibtex for seamless addition into LaTeX / Overleaf (references.bib).
    • synaptik cite --apa for APA 7th edition citation strings.
    • synaptik cite --json for programmatic integration into computational pipelines.
• Citation File Format (CITATION.cff v1.2.0):
  • Formatted with verified author ORCID (https://orcid.org/0009-0001-9477-3332), powering GitHub's native "Cite this repository" right-hand sidebar widget.
• Translational Workflow Citation Footers:
  • Integrated attribution footers into all core analytical outputs (synaptik protocol, synaptik design, synaptik aso, synaptik biomarker, synaptik simulate, and synaptik translate).
• Package Authorship in pyproject.toml:
  • PyPI and wheel package metadata registered with Francesco Nobile as author and maintainer.
• Zenodo DOI Integration:
  • Configured repository architecture for automatic minting of persistent Digital Object Identifiers (DOIs) indexed in Crossref and Google Scholar upon release.
• Verification:
  • 20 new tests in tests/test_cite.py, bringing the total platform test suite to 362 passing checks across 8 test suites on Python 3.10, 3.11, and 3.12 with 100% pass rate.

v0.6.0 (Alpha)

SYNAPTIK v0.6.0: Preclinical Study Design Registry & Multi-Site Meta-Analysis

This release introduces pre-registration grade preclinical study design protocols, randomized complete block plate allocation with edge-effect avoidance, cryptographic SHA-256 protocol fingerprinting, and multi-site inverse-variance random-effects meta-analysis.

Highlights

• synaptik protocol CLI command & API:
  • Generates reproducible, pre-registration grade experimental design protocols compliant with Nature/PLoS reproducibility guidelines (Landis et al. Nature 2012, Macleod et al. PLoS Biol 2015).
  • Pre-specifies primary and secondary endpoints, target disease models, maturity DIV, and paired sample sizes with exact power calculations.
• Cryptographic pre-registration fingerprinting:
  • Produces an immutable SHA-256 design digest and blinding verification token (SYN-...) guaranteeing protocol lock prior to unblinding and analysis.
• Randomized block plate architecture:
  • Automated 96-well allocation with perimeter evaporation barriers (reserving outer wells for PBS) and alternating mutant / isogenic control pairs to eliminate spatial and edge confounds.
• Multi-site random-effects meta-analysis:
  • Synthesizes replication datasets across independent research institutions using the DerSimonian-Laird inverse-variance random-effects model.
  • Reports heterogeneity statistics ($Q, \tau^2, I^2$) and executes automated leave-one-site-out sensitivity testing to ensure pooled conclusions do not depend on a single laboratory.
• Test coverage:
  • 21 new regression checks in tests/test_protocol.py, bringing the total platform test suite to 342 passing checks across Python 3.10, 3.11, and 3.12.

v0.5.1 (Alpha)

SYNAPTIK v0.5.1: Translational Fluid Biomarker & Patient Stratification Panel

This release bridges in vitro disease model findings directly to clinical fluid biomarkers (Neurofilament Light Chain, NfL) and patient cohort stratification for clinical trial optimization.

Highlights

• synaptik biomarker CLI command & API:
  • Models longitudinal in vivo biomarker suppression (plasma and CSF NfL via Simoa ultrasensitive assay) from in vitro axonal rescue percentages.
  • Implements clearance pharmacodynamics ($t_{1/2} = 8\text{ weeks}$, translation scaling factor $\beta = 0.72$), projecting suppression curves at Weeks 4, 8, 12, 16, 24, 36, and 48.
  • Benchmarked against clinical trial precedents: VALOR study (SOD1 tofersen, Miller et al. NEJM 2022: ~60% plasma NfL suppression; FDA Accelerated Approval surrogate endpoint framework).
• Clinical trial patient stratification & cohort optimization:
  • Classifies patients into progression tiers (fast, intermediate, slow) based on baseline plasma NfL cutoffs ($&gt; 75\text{ pg/mL}$, $40\text{--}75\text{ pg/mL}$, $&lt; 40\text{ pg/mL}$).
  • Computes trial power and sample size requirements: demonstrates that biomarker-enriched patient stratification reduces required patient cohort size per arm by 60–75% (e.g. from 74 down to 18 patients/arm) to detect 30% therapeutic slowing of functional decline (ALSFRS-R).
• Test coverage:
  • 21 new regression checks in tests/test_biomarker.py, bringing the total platform test suite to 321 passing checks across Python 3.10, 3.11, and 3.12.

v0.5.0 (Alpha)

SYNAPTIK v0.5.0: Causal Genomic Counterfactual Simulator

This release introduces in silico genetic perturbation modeling to predict how candidate therapeutic interventions reverse diseased neuronal transcriptomic vectors.

Highlights

• synaptik simulate CLI command & API:
  • Simulates in silico gene knockdowns, knockouts, cDNA overexpression, and ASO steric blocking against disease baseline vectors (ALS, FTD, AD).
  • Grounded in human iPSC-neuron functional genomics and Perturb-seq benchmarks (Tian et al. Neuron 2019, Replogle et al. Cell 2022).
• Multi-dimensional rescue metrics:
  • Vector Rescue Efficiency (%): Euclidean norm reduction across diseased transcriptomic vectors.
  • Cosine Directional Alignment ($\cos \theta$): Directional alignment against the inverted disease vector ($\vec{v}{\text{rescue}} \cdot (-\vec{v}{\text{disease}})$).
  • Collateral Toxicity Index: Monitors unintended disruption of neural housekeeping genes (ACTB, GAPDH, TUBB3, SNAP25) and flags pan-essential common lethal genes (POLR2A, RPS6).
• High-throughput target screening:
  • synaptik simulate --disease als --screen ranks cataloged interventions against disease baseline vectors, correctly surfacing viable therapeutic candidates (STMN2 ASO steric block, ATXN2 knockdown, UNC13A ASO block, PIKFYVE knockdown) while rejecting toxic and pan-essential controls (POLR2A, TARDBP overexpression).
• Test coverage:
  • 24 new regression checks in tests/test_simulation.py, bringing the total platform test suite to 300 passing checks across Python 3.10, 3.11, and 3.12.

v0.4.1 (Alpha)

SYNAPTIK v0.4.1: Dish-to-Brain Human Post-Mortem snRNA-seq Bridge

This release connects in vitro findings from human iPSC-derived neuron disease models directly to human post-mortem single-nucleus transcriptomic atlases.

Highlights

• synaptik translate CLI command & API:
  • Compares observed in vitro perturbation profiles (or differential splicing tables from MAJIQ / rMATS / LeafCutter) against hallmark post-mortem human brain pathology.
  • Curated landmark atlases for Amyotrophic Lateral Sclerosis (ALS) (NYGC ALS Consortium), Frontotemporal Dementia (FTD), and Alzheimer's Disease (AD) (SEA-AD).
• Pathology-aware concordance scoring:
  • Differentiates cryptic exon inclusion pathology (where cryptic inclusion $E(\Delta\Psi) &gt; 0$ reflects TDP-43 nuclear loss of function, such as in STMN2, UNC13A, and AARS1) from standard cassette exon skipping.
  • Computes directional concordance percentage, identifies directional contradictions, and renders translational verdicts (HIGH TRANSLATIONAL FIDELITY, MODERATE TRANSLATIONAL FIDELITY, MIXED TRANSLATIONAL FIDELITY, TRANSLATIONAL DISCORDANCE DETECTED).
• Cell-type specificity mapping:
  • Maps dysregulated genes to specific CNS cell subpopulations: Upper Motor Neurons / Betz cells, Lower Motor Neurons / Spinal cord, Cortical Projection Neurons (Layers 2-6), GABAergic Interneurons, Astrocytes, Microglia, and Oligodendrocytes.
• Developmental maturation gap detection:
  • Identifies adult brain maturation markers (NEFH, GRIN2A) that naturally lack expression in early iPSC neurons (DIV 21–35), flagging them as developmental maturation gaps rather than disease discordances and offering actionable culture recommendations.
• Test coverage:
  • 19 new regression tests in tests/test_translation.py, bringing the total platform test suite to 276 passing checks across Python 3.10, 3.11, and 3.12.

v0.4.0

SYNAPTIK v0.4.0: In Silico Antisense Oligonucleotide (ASO) Engineering Engine

This release marks a major translational leap, transforming SYNAPTIK into an active drug-design engine for neurological diseases.

Highlights

• In Silico ASO Engineering Pipeline (\synaptik aso):
  • Automatically engineers clinical-grade antisense oligonucleotides for splice-switching (SSO, steric hindrance for cryptic exon exclusion or exon inclusion) and RNase H1-mediated transcript degradation (gapmers).
  • Nearest-neighbor thermodynamic duplex modeling ($\Delta H^\circ$, $\Delta S^\circ$, $\Delta G_{37}^\circ$, $) under physiological CSF conditions ($[Na^+]=100\text{ mM}, [C_T]=250\text{ nM}$) with empirical 2'-MOE thermal stabilization ($+1.1^\circ\text{C}$/nt), matching clinical empirical data within .6^\circ\text{C}$ (e.g. Nusinersen \approx 64.5^\circ\text{C}$).
  • Biophysical liability filters detecting G-quadruplexes ($\ge 4$ Gs; neurotoxicity risk), poly-T/U tracts, secondary hairpins ($\Delta G_{\text{hairpin}}$), and 3'-end self-dimerization.
  • Neural transcriptome off-target screening against core housekeeping and neuronal genes (ACTB, GAPDH, TUBB3, MAP2, SYP, SNAP25, TARDBP, FUS, SOD1, C9orf72), enforcing zero 0-mismatch or 1-mismatch off-target hits.
  • Synthesis-ready output: Molecular weight (sodium salt), extinction coefficient $\epsilon_{260}$, predicted human CSF terminal half-life (~135 days), and orderable IDT vendor format strings (\mN*...).
• Comprehensive Quality Assurance:
  • Dedicated test suite \ ests/test_aso.py\ with 20 new checks (257 checks total across the platform).
  • 100% CI pass rate across Python 3.10, 3.11, and 3.12.

v0.3.1

SYNAPTIK v0.3.1

Highlights

• Native MAJIQ & LeafCutter Splicing Ingestion: Direct auto-detection of MAJIQ (VOILA) and LeafCutter tables in synaptik deposit splicing, parsing multi-junction LSVs and cluster coordinates for reading-frame and NMD consequence evaluation.
• Paired Isogenic Power Analysis: Exact noncentral-\$ paired sample size calculation ( = d / \sqrt{2(1 - \rho)}$) for human iPSC disease models, rescuing prohibitive unpaired differentiations ( > 25-50$) down to feasible paired cohorts ( = 6-10$ pairs). Configurable via synaptik design --rho .
• Neuronal Splice Prediction Benchmark Suite: Gold-standard validation suite (synaptik benchmark --spliceai) testing sequence models against published neurodegenerative splice variants (STMN2, UNC13A, SMN2, ELP1, COL6A1) and matched neutral controls (Sensitivity: 100%, Specificity: 100%, ROC-AUC: 1.000).
• Comprehensive Quality Assurance: 10 new regression checks added (237 checks total across the suite); 100% test pass rate verified on GitHub Actions CI across Python 3.10, 3.11, and 3.12.