Illustrative model example
A modification can change the charge premise before the design changes.
Take a synthetic 36-residue sequence containing three blocks: 12 lysines, 12 serines and 12 alanines. Replacing the serines with glutamate as a phosphorylation mimic changes both the charged fraction and net charge. That gives the next investigation a concrete starting point: compare the relevant protein states before carrying forward a sequence-based intervention hypothesis.
- K₁₂ · S₁₂ · A₁₂Starting sequence
Charged fraction: 1/3. Net charge per residue: +1/3.
- S → EApply the charge mimic
Replace the 12 serines with glutamate in the model.
- K₁₂ · E₁₂ · A₁₂Recalculate the premise
Charged fraction: 2/3. Net charge per residue: 0.
- What changes · The original sequence has 12 positive and no negative residues. The mimic has 12 of each: twice as many charged residues, with balanced net charge. Charge patterning can then be recalculated on that explicit sequence.
- What the boundary check adds · BioTwin also checks whether a supplied modification annotated as disease-causal lies near the proposed intervention boundary. The causal annotation is an input to the check. That connects the sequence change to a specific reason to revisit the hypothesis. The numerical example does not establish a changed binding site or a successful intervention.
- The decision it supports · Choose a comparison of unmodified and authentically modified protein states before investing in a design that assumes they are equivalent. localCIDER already supports charge and phosphorylation-mimic analysis; the value here is carrying the changed premise into the intervention decision.
Recalculate the premise
The methods examine charge patterning, hydropathy, charged-residue fraction and net charge. Modification sites near a proposed boundary are checked, and sequence priors can be recalculated using explicit mimetic substitutions. The question is whether the same intervention hypothesis still makes sense.
Put the changed sequence back into the design
Tools such as localCIDER already calculate sequence parameters and simplified phosphorylation effects. BioTwin’s intended contribution is to carry the changed sequence context into the boundary or intervention decision being investigated. A predictive advantage over specialist IDP methods has not been demonstrated.
A sequence calculation sets up the experiment
Current methods generate hypotheses; they do not establish a binding site or druggability. Modification mimics reproduce only part of the chemistry. Compare the relevant states using a measurement sensitive to disorder, preferably including the authentic modification.
Check which protein product the result belongs to
Isoform differences and interface evidence can change the interpretation again. These are useful adjacent investigations, with the combined analysis qualified for the particular target.
- Isoform context · Separate products with different sequence or function. Isoform context
- Interface constraints · Test a proposed interaction against the relevant state. Interface constraints