ProteinIQ for protein design

Protein design that actually works

Generate optimized protein sequences and novel binders using LigandMPNN, ProteinMPNN, SolubleMPNN, BoltzGen, and RFdiffusion—state-of-the-art deep learning and diffusion models.

BindCraft

End-to-end binder design pipeline combining RFdiffusion and AlphaFold2. Design high-affinity binders with experimental validation.

ProteinMPNN

State-of-the-art inverse folding with >50% sequence recovery. Design sequences for any backbone structure.

LigandMPNN

Design sequences with ligand, metal, and nucleotide context. 63.3% recovery at binding sites.

SolubleMPNN

Design soluble protein sequences optimized for expression and stability. Reduce aggregation risks.

EvoDiff

Generate diverse protein sequences using evolutionary-scale diffusion models. Unconditional and conditional generation.

ESM-IF1

Inverse folding with ESM language models. Fast sequence design from structure with confidence scores.

Why researchers choose ProteinIQ

AI-guided design generates optimized sequences in minutes

Directed evolution requires months of experimental iteration

BoltzGen and RFdiffusion generate binders with predicted nanomolar affinity

Designing binders for novel targets is hit-or-miss

Upload structure, configure options, get sequences—no coding required

Inverse folding tools have steep learning curves

SolubleMPNN optimizes for solubility from the start

Solubility and expression issues discovered late in development

How it works

Design protein sequences and structures using cutting-edge deep learning and diffusion models. Inverse folding with LigandMPNN, ProteinMPNN, and SolubleMPNN, plus binder and structure generation with BoltzGen and RFdiffusion.

Upload structure

Provide target backbone or binding target PDB

Configure design

Set residue constraints, chain specifications

Generate sequences

ProteinMPNN or LigandMPNN designs optimized sequences

Validate & order

Check predictions and export for synthesis