Repository navigation
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 simulateCLI 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 --screenranks 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.
- 24 new regression checks in