- stable structural core
- persistent contacts
- defined interaction surfaces
- structure often explains function
EWCL
Entropy-Weighted Collapse Likelihood
Interpretable residue-level prediction of protein disorder and collapse propensity.
Analyze how sequence physics, structural confidence, and local geometry shape protein disorder and conformational behavior.
Explore example analyses
Why disorder matters
Proteins do not always function through one stable structure.
Intrinsically disordered proteins and regions remain dynamic, sampling many conformations rather than settling into a single fold. This flexibility supports signaling, molecular recognition, binding-induced folding, regulation, and biomolecular condensates.
- conformational ensemble
- transient and context-dependent contacts
- changing interaction surfaces
- sequence grammar and ensemble behavior matter
What EWCL measures
Sequence physics becomes residue-level evidence.
EWCL converts explicit sequence-derived physicochemical properties into a continuous residue-level disorder profile. Each prediction can be examined through the physical feature families that support it and, when structural data are available, compared directly with AlphaFold confidence and local geometry.
In EWCL, entropy and collapse refer to sequence-derived descriptors associated with compositional diversity, local conformational freedom, interaction balance, and compaction propensity. The final score is a supervised disorder likelihood.
Feature space projected into an interpretable residue profile.
Sequence-derived features
Amino-acid composition, charge patterning, hydropathy, local complexity, flexibility, entropy-related descriptors, interaction balance, and collapse propensity.
Residue-level prediction
A continuous score estimates whether the local sequence environment is more consistent with structural compaction or conformational heterogeneity.
Feature-family interpretation
Predictions can be traced to contributing physical feature families instead of being presented only as a final probability.
Structure-aware comparison
AlphaFold pLDDT and local geometric descriptors are used to examine agreement, uncertainty, and Confidence-Disorder Conflict regions.
Analysis preview
One score, multiple levels of interpretation.
The analysis view keeps the continuous profile visible while linking local regions to evidence overlays, AlphaFold comparison, Confidence-Disorder Conflicts, and feature-family explanations.
Example output
P04637 | p53 residue analysis
Feature-family heatmap
blue order-supporting | red disorder-supportingResidues 64-92 show elevated sequence-derived disorder propensity together with low structural confidence, supporting a conventional disordered-region interpretation.
Residues 173-188 show elevated EWCL disorder propensity despite high pLDDT and are marked as a Confidence-Disorder Conflict region for further inspection.