The Decision Layer for Protein R&D

Better questions Better workflows Better decisions

An agentic integration & decision platform for explainable protein discovery, design, and optimization, from targets to candidates.

Request Access See How It Works
ProtXplain, protein R&D platform
ProtXplore
ProtFlow
I want to design a novel antibody against human PD-L1.
PD-L1 is a validated oncology drug target. I retrieved its 3D protein structure and identified a PD-L2-specific binding hotspot. Shall we start with designing a specific antibody?

The Challenge

Protein R&D is fragmented.
Decision-making has become the bottleneck.

Researchers waste weeks stitching together incompatible tools, chasing unreliable results, and re-running analyses from scratch. ProtXplain changes that.

Siloed Tools, Siloed Insights

Structure prediction, docking, and sequence analysis live in separate universes, forcing manual handoffs and error-prone copy-paste workflows.

Decisions Without Context

Critical choices, which target to pursue, which variant to test, are made without a unified view of the evidence, leading to costly late-stage failures.

Irreproducible Workflows

Ad-hoc scripts and undocumented pipelines make it nearly impossible to reproduce, audit, or build upon prior work, even within the same team.

The Solution

One integrated decision layer.
Structured. Explainable. Actionable.

ProtXplain combines an intelligent research agent with a visual workflow engine, giving your team both depth of insight and control, and turning fragmented analyses into clear, auditable decisions.

Research Agent

ProtXplore

A conversational AI agent with deep protein knowledge. Ask questions in natural language, retrieve structures, identify binding hotspots, and generate data-driven hypotheses, all in one place.

  • Natural language queries over PDB, UniProt & AlphaFold DB
  • Automated binding site & hotspot identification
  • Literature-grounded answers with citations
  • Seamless handoff to ProtFlow for execution
Workflow Engine

ProtFlow

A visual, node-based workflow builder for protein R&D. Connect AlphaFold, ProteinMPNN, ESM-2, and custom tools into reproducible, auditable pipelines, without writing a single line of code.

  • Drag-and-drop workflow builder with 100+ nodes and agent-assisted pipeline generation
  • Native integrations with leading protein AI tools such as AlphaFold, ESM-2, and ProteinMPNN
  • Ready-to-use templates for antibody design and more
  • Full execution logs and decision traceability

Workflow

From questions to decisions.

ProtXplain connects the entire protein R&D loop, from exploratory questions to structured, reproducible experiments.

01

Ask ProtXplore

Start with a natural language question. ProtXplore retrieves relevant structures, surfaces binding insights, and generates data-driven hypotheses grounded in literature and data.

02

Build in ProtFlow

Convert your hypothesis into an executable workflow. Drag, connect, and configure nodes for folding, design, docking, and validation, without writing code.

03

Decide with confidence

Get structured, auditable results. Compare variants, track decisions, and share reproducible workflows with your team and collaborators.

Use Cases

Built for real-world
protein decisions.

From academic labs to biotech pipelines, ProtXplain accelerates decision-making across the most critical protein R&D workflows.

01

Candidate Discovery & Functional Similarity

Identify novel candidates and functional similarities across proteins, uncovering hidden relationships and better starting points for discovery and design.

02

Explainable Protein Function Prediction

Predict and rank protein function with interpretable, auditable reasoning. Identify substrate specificity and functional signals, even in low-data settings.

03

Antibody & Binder Design

Design and prioritize antibodies and protein binders. Identify epitopes, generate candidates, and rank variants by affinity, specificity, and developability, from target to decision.

04

Enzyme Engineering

Design and optimize enzymes by screening and ranking mutations for activity, stability, and selectivity, prioritizing top variants before wet-lab validation.

05

Proteomics & Biomarker Discovery

Rank biomarkers across datasets and cohorts with high confidence. Generate reproducible, traceable results with clear biological and clinical relevance.

06

Target Validation & Biosafety Assessment

Evaluate and prioritize targets based on structural context, druggability, and biological evidence, with early toxicity detection enabling safer go/no-go decisions before design.

The Team

Science meets
engineering.

A founding team combining deep computational biology, AI engineering, and biotech experience.

Founding Team

Stephan Breimann
CEO

Stephan Breimann

PhD in Computational Biology (TUM). First-author publication in Nature Communications (2025). Co-founder of CORE64.

Protein Design Explainable AI Proteomics
Ivan Koludarov
CSO

Ivan Koludarov

PhD Molecular Evolution (U. Queensland). 60+ publications, including Nature Genetics (2020) and Nature Communications (2023).

Protein LMs Evolutionary Genomics Bioinformatics
Mohammad Amin Sohrabi
CPO

Mohammad Amin Sohrabi

MSc CS (AI), PhD candidate Computational Biology (TUM). Bridges agentic AI engineering and product design.

AI Full-Stack UI/UX

Development & Application

Peyman Vahidi
Agentic AI Engineer

Peyman Vahidi

PhD candidate at Helmholtz Munich. Brings agentic AI systems expertise to ProtXplain's orchestration layer.

Agentic AI Bioinformatics
Tobias Senoner
Computational Biologist

Tobias Senoner

PhD candidate at TUM. Applies computational methods to protein sequence and structure analysis across the ProtXplain pipeline.

Computational Biology Protein LMs
Florin Senoner
Software Engineer

Florin Senoner

ETH Zürich alum. Full-stack engineer building the platform infrastructure, frontend, and integrations that power the ProtXplain experience.

Full-Stack Platform Engineering
Srikanth Lingappa
Protein Scientist

Srikanth Lingappa

PhD from the University of Bristol. Bridges computational protein design with wet-lab validation to ground ProtXplain's predictions in experimental reality.

Protein Design Wet-Lab Validation

Advisory Board

Prof. Thomas Schlichthärle
Protein AI Advisor

Prof. Thomas Schlichthärle

AI-driven Protein Design · Ex-Baker Lab

Expert in AI-driven protein design, formerly at the Baker Lab (University of Washington) — a world-leading group in computational protein design and structure prediction.

Prof. Dmitrij Frishman
Bioinformatics Advisor

Prof. Dmitrij Frishman

Bioinformatics · Co-founder, BioMax

Professor of bioinformatics and co-founder of BioMax, bringing deep expertise in computational biology, sequence analysis, and protein function prediction.

Early Access

Ready to transform
how you make decisions?

Join the researchers and biotech teams already using ProtXplain to make faster, better-grounded decisions in protein design.

Request Early Access Talk to us