Welcome to InsilicoΣ Lab

Build predictive QSAR models using 2D, 3D, and 4D molecular descriptors with machine learning algorithms.

Predict ADMET properties, pharmacokinetics, and drug-likeness to evaluate compound safety and efficacy.

Predict polymer properties (Tg, Tm, density, solubility) using 3-tier prediction: group contribution, ML ensemble, and polyBERT.

Generate 256-bit bio-context fingerprints for drug targets, encoding membrane topology, signaling, binding pocket, and endogenous ligands.

Design siRNA, sgRNA (CRISPR), ASO/GapmeR, and optimized mRNA with integrated SafeRNA immunogenicity profiling and off-target analysis.

Design and analyze polymer structures with automated polymer builder and property analysis tools.

Analyze drug-target-disease relationships through network pharmacology and systems biology approaches.

User-friendly machine learning for clinical data analysis. Apply ML methods to patient data, biomarkers, and clinical outcomes without coding.

Conduct full meta-analyses with forest plots, effect size calculations, heterogeneity analysis, and publication bias assessment.

Assess risk of bias using 7 validated instruments (RoB 2, ROBINS-I, NOS, QUADAS-2, QUIPS, AMSTAR 2) with guided domain-by-domain wizard and automated judgment algorithms.

AI-augmented GRADE evidence evaluation with rule-based domain assessment, OIS calculation, Summary of Findings tables, and publication-ready Word/PDF export.

Standardize extracted study data before pooling: convert measurement units (mg/dL ↔ mmol/L, 13 analytes) and derive SD from any reported dispersion measure (SEM, 95% CI, IQR, range) using Wan 2014 methods.

Generate novel drug candidates using reinforcement learning and generative AI for de novo molecular design.

Generate 3D drug molecules inside protein binding pockets using decomposed diffusion with scaffold-arm decomposition.

Design protein binders (nanobodies, peptides, mini-proteins) for any target using Boltzmann-distribution generative models.

Predict 3D protein structures from amino acid sequences with per-residue confidence scores and interactive 3D visualization.