We are innovating in the protein design field. Our business activity will start in 2023.

By combining physics, robotics-inspired algorithms, machine learning and automated reasoning, we design improved proteins from first principles and provide customers with optimized molecules for all type of applications.

Time saving

Cost reduction


Success rate

Time and cost saving

Our pure in silico design processes avoid most of the expensive and time consuming experimental essays. Only carefully selected designs need to be tested.

Beyond natural proteins

Our AI-powered design method is able to confer new, desirable properties, providing sequences that evolution has not been able to explore.

Customized solutions

With its extreme flexibility, our design engine can offer customized solutions that match your needs: specificity, stability, sequence composition, activity,...

A palette of services

Every design problem is different. We offer a variety of services, adapted to a range of different requirements, from the simplest to the most challenging ones.


A team of dedicated experienced professionals.

Sophie Barbe



Sophie is a Research Director at INRAE, specialized in protein modeling and design. She has a long experience of interactions with academic labs and companies for the tailor-made conception of proteins with improved and new capabilities for Biotechnology and Health. Her expertise covers computational protein design, methodological developments for structural bioinformatics, structure-function relationship studies, and rational protein engineering. She has co-authored over 60 scientific papers and 5 patents and has contributed to the development of several molecular and design methods (AI-based).

Juan Cortes



Juan is a Research Director at CNRS and a computer scientist specialized in algorithmic robotics. He is one of the pioneers in the development of structural bioinformatics approaches based on algorithms originating from robotics and artificial intelligence. During the last 15 years, he has conducted fruitful interdisciplinary research in this area in collaboration with bio-chemists, biologists and physicists. He has co-authored over 90 scientific papers and coordinated the development of several software packages and tools, in particular MoMA.

Thomas Schiex



Thomas is a Research director at INRAE, a computer scientist specialized in hybrid systems combining logical and numerical automated reasoning with probabilistic machine learning and deep learning methods in AI. He also contributed to their extension and applications in computational biology from genetic, genomics to structural biology. He has co-authored more than 160 scientific papers and several original bioinformatics and core-AI software. He is a Fellow of the Association for the Advancement of AI and of the European Association for AI.


Characterized proteins designed with Amineo's technology

A new nanobody with unique sequence composition, expressibility and genericity for bio-diagnosis

In collaboration with the CRCT (Centre de Recherches contre le Cancer de Toulouse) with funding from INSERM Transfert, our AI technology was able to design a completely new nanobody scaffold. As targeted, the designed nanobody has high expressibility, unique fluorescence capacities for in-cell diagnosis and keeps affinity for targets with various sets of CDR loops (patent pending). A result of our hybrid AI multi-state protein design capacities.

Improving activity, thermo stability and resistance: making your enzyme tougher.

GH-11 xylanases are able to break the long chains of xylans, a major components of available biomass. They have a wide range of applications, in bioenergy, food and feed and paper industry. We simultaneously improved the activity, thermo-stability and heat-shock resistance of one of the most active, already engineered, GH-11 enzyme (patent pending). Another result of our advanced AI multi-state design capabilities.

A self-assembling protein that resist extreme pH and temperatures.

Previous attempts at designing symmetric beta-propellers had a hard time to reach a stable design because of the unique strain that beta-propellers backbone must cope with. Thanks to our guaranteed automated reasoning design process, in collaboration with KU-Leuven and Riken, we were able to design a new hyper-stable self-assembling beta propeller that can resist extreme conditions in terms of pH and temperature. Our first proof of concept, now published in the IUCrJ journal.

We don't trace you. This website is using functionnal cookies only. That's Fine