Neuralgap

Neuralgap is building interpretable deep learning for structure-based molecular discovery, connecting predictions to mechanistic evidence with calibrated uncertainty and explicit attribution.

Interpretable Computational Biophysics

Read more

Two products on the discovery path: Bioforager for chemotype funneling, Spyndle for structure-guided design.

The Funnel Products

See Bioforager See Spyndle
Bioforager: ligand interaction analysis
Bioforager

Ensemble AI and physics-based screening for peptides and ligands.

Spyndle dashboard: loop design workspace
Spyndle

Mechanistic triage for delivery: AAV now, LNPs next.

What we cover

Platform coverage

Explainability-guided triage across two fields: peptides and ligands, and delivery vehicles.

Peptides & ligands

  • Ensemble AI and physics-based virtual screening
  • Peptide and ligand optimisation, including organometallics
  • Undruggable and structure-enabled targets
  • Interpretable gates before synthesis

Enabled by Bioforager

Delivery

  • Mechanistic triage for delivery vehicles
  • AAV capsid design, in use now
  • LNPs, forthcoming
  • Structure-guided loop and payload decisions

Enabled by Spyndle

Shared internals: ensembles, QM Hamiltonians, pathway-aware MD, and attribution you can inspect.

Clinical and discovery programmes

Pipeline

Oak Crest Institute of Science

SAL 001-004 · Early stage

SAL-001

Early discovery

SAL-002

Early discovery

SAL-003

Early discovery

SAL-004

Early discovery

Oncology

C-Myc

Oncology target

About Neuralgap

Seeing into structure, not past it

Interpretable molecular discovery

We build models for structure-based drug discovery that keep mechanistic evidence in the loop: conformational ensembles, homology-aware selectivity, and attribution you can inspect.

The aim is simple. When a candidate is promoted or killed, a chemist should be able to tell model artefact from biological signal, and say why.

Why Neuralgap?

Neuralgap is a computational biophysics studio: explainability-guided triage for undruggable targets, and for delivery (AAV capsids now, LNPs next) so a design can be trusted, or set aside, with a reason.

This is Neuralgap

Meet the team

Decades of experience in explainable AI and biochemistry research for structure-based drug design.

Gamika Seneviratne

Gamika Seneviratne, MEng

Co-Founder · Chief Executive Officer

Serial entrepreneur who pioneered explainability algorithms including Granger Causality and regime-switching methods, with direct application in drug design. MEng, Imperial College London.

Juliyan Gunasinghe

Juliyan Gunasinghe

Co-Founder · Chief Scientific Officer

Cheminformatics and medicinal chemistry researcher pursuing a PhD at the University of Melbourne. Published in mechanism-encoded SAR, chemical space topology, and ligand optimisation.

Sathursan Kanagarajah

Sathursan Kanagarajah

Co-Founder · Chief Technical Officer

Serial entrepreneur and senior AI engineer, with computer vision research at NTU Singapore. Pioneering work in dynamic explainable representations; published at ICCV and ISBI.

Manula Vishvajith

Manula Vishvajith

Co-Founder · Chief Backend Designer

Former PwC MLOps Senior Lead Architect and ISO 27001 certified auditor. Designs agentic AI systems and production-grade platforms that underpin Neuralgap’s discovery workflows.

Selected papers

Publications