Page progress

Our Technology

Two complementary AI platforms working together to revolutionize genetic medicine

Suppressor tRNA Design Engine

The Suppressor tRNA Engine decomposes tRNA into functional blocks (acceptor stem, D-arm, anticodon loop, variable loop, TΨC loop, CCA) and proposes candidate suppressor tRNAs for a given gene + PTC. Hard filters enforce baseline tRNA viability and recognition (tRNAscan-SE, SPOT-RNA2/RNAformer structure checks, CCA/RNase P site integrity, aaRS identity elements). Scoring integrates: ΔG folding (ViennaRNA/EternaFold), orthogonality heuristics, similarity to known suppressors, MD fluctuation (OpenMM backend, RMSF per domain), and a cis-context penalty derived from real termination annotations (TransTermHP/APPRIS). Candidates loop through rank → wet-lab validation → model updates, targeting India-first clinical translation.

tRNA Design Engine Pipeline

AI-driven suppressor tRNA design and validation workflow

Stage 1 of 10
01

Input

Gene + PTC position

02

Generation

Candidate tRNAs

03

Hard Filters

Viability checks

04

Thermodynamics

ΔG folding

05

MD Simulation

RMSF analysis

06

Similarity

Known suppressors

07

Cis-Context

Local penalties

08

Ranking

Score & explain

09

Pre-clinical

Wet-lab validation

10

Clinical

Translation ready

Processing: Input
Primary Flow
Filtering
Analysis
Feedback Loop
KritRNA Rescuing Nonsense Mutations via Suppressor tRNA

AI-Powered Design

Machine learning models predict optimal suppressor tRNA sequences

Context Awareness

Considers local sequence context and cellular environment

Efficiency Optimization

Maximizes suppression while minimizing off-target effects

Notes: Nonsense/PTC share ≈10–11% across diseases; TP53 truncations are frequent in cancer. India context follows NPRD 2021 policy with 1,118 beneficiaries reported (Aug 9, 2024).