Kairos / PichiaCandidate compression

Which Pichia constructshould be tested first?

Enter a target protein and process constraints to receive 10–20 constructs ranked by evidence and risk.

Dry-lab design ranking · no absolute titer prediction

01 / The new bottleneck

Candidate sequences far outnumber verifiable wells

Wet-lab capacity, expert attention and defensible failure analysis are scarce.

More candidates do not choose the next experiment

Kairos combines host, construct, evidence and process constraints in one trace.

Before
Many candidates
During
Evidence review
After
Ranked assay plan

02 / Canonical public record

From candidate to experiment: a traceable decision record

Cross-species and copy-number heuristics stay visible as hypotheses.

Ubiquitin secretion in K. phaffii

Can α-mating factor direct ubiquitin into secretion? The record keeps its ranked CDS candidates and unresolved boundaries visible.

Record
—
Gate
Provisional
Sources
—
Findings
—
Candidates
—
Date
2026-11-16
Failure modeHeuristic retained

03 / What the workbench returns

Rank candidate constructs by evidence and risk

A defensible next-round order, with evidence and open questions.

01

Candidate compression

10–20 constructs worth a limited wet-lab slot.

What to build
02

Risk compilation

Seven axes of likely failure, with conflicts kept visible.

Where it fails
03

Evidence audit

Supported, conflicted and unverified claims stay distinct.

What is defensible

04 / Scope and fixed-scope pilot

Capability boundary: decisions before wet-lab entry

No titer promise, wet-lab replacement, FTO advice or unvalidated generalization.

Read the full capability boundary
Host
Pichia only
Evidence
Directional, partly provisional
Next gate
Prospective wet-lab readout
Not claimed
Expression, recovery, scale-up or release

Start with one real target

Inputs, deliverables and acceptance criteria are fixed in writing.