Industrial design intake

Describe the process. We'll work out the molecule.

You don't need to know which model to run, or what a diffusion backbone is. Tell us what the enzyme has to survive and what it has to cost.

Start from a process problem

Pick a bottleneck to load, or describe your own below.

Design brief

Plain language. Edit anything — this is what the intake model reads.

Your process window

The conditions the molecule has to survive. These drive the design objective, not the other way round.

70 °CpH 4.015% solids2 g/L Phenolics48 hP. pastoris$8/kg
First-pass design objective
  • Hold ≥50% activity for 48 h at 70 °C — implies Tm ≥ 88 °C
  • Acid-stable at pH 4 — redesign surface charge to suppress low-pH unfolding
  • 2 g/L Phenolics present — screen designs for inhibitor binding at the active site
  • Stable at 15% solids — resist shear and surface adsorption losses
  • Secrete in P. pastoris at a titre consistent with $8/kg product

Rule-based first pass. The intake model refines this against the knowledge base once you continue.

How would you solve this today?

Pick your real alternative. Every number in this session is compared against it.

Designing this pipeline by hand
Experienced protein-design scientist

Knows the model landscape. Reads the target, picks the stack, sets contigs and hotspots, writes the notebook.

2–3 days
Biochemist without a design background

Has to learn the landscape first — and still lands on a weaker stack than the specialist would.

3–6 weeks
Chondrule

Assembled against the knowledge base, with the reasoning and the rejected routes shown.

under 2 minutes

This is not only elapsed time. The choice of stack is what sets the wet-lab hit rate — a weaker pipeline fails at the bench, months later, not on screen.

Comparing against CRO / enzyme-major campaign·6–9 months·licence fee on output

Next: knowledge-base-driven PDB search, then hypotheses and pipeline assembly.

or