Computational biologic design & engineering

Your next biologic
starts here.

Put AI and in silico tools to work on your next peptide, miniprotein, VHH, or antibody.

From your target to a focused set of designs: generate binders, engineer the full molecule, and assess developability—with a clear plan for experimental testing.

Start your design project
Design starts at the interface.

Loading antibody–antigen structure…

FabTarget

Drag to explore · Scroll to zoom

Eduard Puig / Ph.D.Based in San Diego · Working globally

01 / Expertise

Design. Assess. Optimize.
Build the tools behind it.

Bring a target, a binder, or a discovery bottleneck. Get computational designs, experimental guidance, or custom software matched to your program.

01 / DESIGN

Molecular design

Design the binder. Engineer the complete molecule.

Design peptides, miniproteins, VHHs, and antibodies de novo. Engineer their format, linkers, and domain geometry with expression, purification, and manufacturing in mind.

Explore capabilities
  • De novo binder design — sequence generation and candidate prioritization across peptides, miniproteins, VHHs, and antibodies.
  • Molecule & modality engineering — binder assembly, domain arrangement, linker geometry, valency, and fusion or multispecific architectures.
  • Structure prediction — modeling individual molecules and complexes to assess interfaces, geometry, and design hypotheses.

Typical deliverables Prioritized sequences, structural models, and construct designs ready to take into experimental testing.

02 / ASSESS & TEST

Developability & validation

Find the risks before choosing what to build.

Screen for molecular liabilities in silico, then plan expression, purification, and characterization. Design and interpret binding assays to understand how your molecule performs.

Explore capabilities
  • Developability screening — hydrophobic and charged patches, aggregation and solubility risk, sequence liabilities, stability indicators, and predicted immunogenicity risk.
  • Experimental guidance — expression-system and construct strategy, purification planning, and biophysical characterization.
  • Binding kinetics — assay design, controls, fitting strategy, and interpretation of affinity, association, and dissociation data.

Typical deliverables A prioritized risk assessment, experimental recommendations, and interpretation of the resulting data.

03 / OPTIMIZE

Hit-to-lead optimization

Improve affinity. Preserve developability.

Focus each optimization round with in silico affinity maturation, targeted yeast-display libraries, and machine learning trained on your experimental data.

Explore capabilities
  • Computational maturation — propose and prioritize variants while balancing affinity, specificity, and developability.
  • Yeast-display guidance — library design, selection strategy, and interpretation of enrichment and sequencing data.
  • Learning from experiments — sequence–function modeling and machine-learning-guided selection of the next variants to test.

Typical deliverables Focused variant libraries, selection plans, and data-informed recommendations for the next optimization cycle.

04 / BUILD

Scientific software & AI

Put computational discovery into daily practice.

Connect your data, models, and experiments with custom software and scientific AI agents. Build workflows your team can run in its own environment, including on premises.

Explore capabilities
  • Scientific agents & workflows — task-specific agents, tool integrations, and automation with appropriate scientist review.
  • Data infrastructure — connect sequence, structure, assay, and project data in usable databases and analysis pipelines.
  • Custom implementation — scientific applications and workflow development, including on-premises deployment, documentation, and team handoff.

Typical deliverables Implemented software and workflows, connected data, and documentation your team can use.

02 / Selected experience

Computational expertise.
Experimental experience.

Experience designing therapeutic molecules, developing assays, and delivering scientific software.

Kelp Bio / Co-founder & lead developer

From peptide data
to a working platform.

Built PepSAR, a peptide structure–activity relationship platform combining chemistry-aware analysis, machine learning, and interactive scientific software.

Commercially licensed to a clinical-stage oncology company.

Valora Therapeutics / Principal scientist

Design with experimental
decisions in mind.

Generated antibody, VHH, and miniprotein candidates; combined structure prediction and interface analysis with experimental triage and developability work.

Experience across bispecifics, cell engagers, humanization, and Fc engineering.

Vertex Pharmaceuticals / Protein science

Make the experiment
more informative.

Built and optimized a high-throughput reporter assay and developed analysis pipelines to investigate compound activity and off-target effects.

Reduced screening cost by approximately 10× in the assay developed.

Prior professional experience; these organizations are not presented as consulting clients.

03 / Working together

One project.
Or part of your team.

Get a focused review, commission a design project, or add ongoing computational expertise to your team. Each engagement starts with clear objectives, deliverables, and decision criteria.

Discuss your project
Eduard Puig, founder of paratope bio

04 / Your scientific partner

Eduard Puig

Pharmacist · Ph.D. San Diego, California

Computational design grounded in protein science.

Eduard Puig is a pharmacist and Ph.D. specializing in computational protein engineering. He helps biotech teams design and de-risk protein therapeutics by connecting in silico design with experimental validation. His consulting spans de novo antibody, VHH, and miniprotein design; affinity maturation and developability; Fc engineering and effector function; and integrated computational–experimental workflows.

Eduard is co-founder of Kelp Bio, where he develops software for exploring peptide and protein structure–activity relationships through molecular analysis and machine learning. As Principal Scientist, Antibody & Protein Design at Valora Therapeutics, he designs de novo binders using tools including RFdiffusion, ProteinMPNN, AlphaFold, ESMFold, and Rosetta, and advances candidates through biophysical and cellular validation.

Previously, as a Fellow at Vertex Pharmaceuticals, he applied structural biology and protein engineering to characterize new CFTR binding sites and developed computational workflows for drug discovery. He held an EMBO postdoctoral fellowship at Scripps Research with Lars Hangartner and Dennis Burton, studying Fcγ receptor biology, HIV Env, and humoral responses to SARS-CoV-2. His doctoral research at IRB Barcelona and IECB focused on amyloid-β oligomer structures and single-domain antibody binders.

Ph.D. in BiochemistryUniversity of Barcelona
EMBO Postdoctoral FellowScripps Research
View professional background Selected publications

05 / Selected publications

The science behind
the practice.

Full publication list on Google Scholar

Start your next biologic

Your target.
Let’s design what binds it.

Tell me about your target, modality, or current bottleneck. We’ll define where computational design can help and what a focused first project could deliver.

Start a conversation