Chain mapping
Heavy and light Dataset columns
Map the paired-chain fields used for every candidate.
Compare predicted antibody viscosity
Classify paired heavy/light candidates with DeepViscosity and compare their predicted high-concentration viscosity risk.
Prepare
Paired antibody sequences
Use a protein or annotated-sequence Dataset with heavy- and light-chain columns on every selected row.
- Paired heavy and light protein sequences per row
- One frozen Selection or Dataset Version
- Up to 100 rows and 500 residues per sequence
Heavy-chain column
heavy_chainLight-chain column
light_chainHow the analysis starts
Choose a Project, open a compatible Dataset, then select the rows you want to analyze. Ubi will open this method with the compatible controls and column mappings ready to review.
Configure the scientific method
These controls appear in the Dataset analysis panel, where values can be checked against the actual input before the run starts.
Candidate label
Optional name column
Carry the candidate name into result tables and comparisons.
Model
DeepViscosity ensemble
Use the same ensemble for every candidate in the comparison.
Review results as scientific outputs
Results open with the figures, structures, sequences, and metrics needed to answer the scientific question. Downloadable files remain available for downstream analysis.
How to interpret the result
- Use the classification to prioritize review within a consistently prepared candidate set.
- The probability is model evidence, not a measured viscosity or a universal formulation threshold.
- Compare predictions with concentration, buffer, temperature, and experimental viscosity data.
1
Candidate classification
Compare predicted viscosity class across the selected antibody set.
2
Ensemble probability
Review mean probability and variation across the model ensemble.
3
Model input features
Inspect the physicochemical feature values used for each candidate.
Continue your analysis
Use a complementary method on the same Project data.
Method scope and limitations
- DeepViscosity requires paired heavy/light chains and does not infer missing partners.
- Predictions support candidate prioritization and do not replace formulation experiments.
