Quartz 4

❯

❯

❯

Design validation

Design validation

Created Sep 26, 2026Modified Sep 07, 2026

Whether computational design scores, refolding tests, or rankings predict useful experimental outcomes; includes appropriate prospective and baseline comparisons.

48 items with this tag.

Sep 07, 2026
ESMFold pLDDT weakly correlates with intra-family differences in stability and experimental success, but not differences in cooperativity
Aug 25, 2026
Ensembling structure prediction methods for filtering improves recovery
Aug 25, 2026
Computational antibody design is more reliable when HCDR3 is held fixed
- design/antibodies
- evidence/design-validation
Aug 25, 2026
Computational ranking of HCDR3 clones can underperform random selection
- evidence/design-validation
- design/antibodies
Aug 25, 2026
Ligand-mimicry filtering improves epitope-targeted antibody discovery
- design/antibodies
- evidence/design-validation
Aug 19, 2026
Confidence metrics are not general binding-affinity predictors
Jul 20, 2026
Experimentally successful miniprotein binders have higher predicted stability
Jul 20, 2026
Most computational protein-stability metrics correlate poorly with experimental dG
- prediction/stability-expression
- evidence/design-validation
Jul 17, 2026
Effectiveness of filtering metrics for de novo minibinder design vary by target
- evidence/design-validation
- design/binders
Jul 17, 2026
Language models cannot extrapolate to functional novelty or ultra-high-fitness variants
Apr 21, 2026
No one-size-fits-all approach to scoring antibody structures
Apr 21, 2026
Adding noise while training non-Ab inverse folding models improves self-consistency while worsening sequence recovery
Apr 21, 2026
AlphaFold2 outperforms ESMFold at distinguishing designable and undesignable protein backbones
- evidence/design-validation
- design/backbones
Apr 21, 2026
Ancestral sequence reconstruction outperforms generative ML methods on functional enzyme design
Apr 21, 2026
Backbone diffusion works best on small protein sizes
- design/backbones
- evidence/design-validation
Apr 21, 2026
De novo designed proteins with alpha helices are easier to predict than those with other secondary structures
Apr 21, 2026
Hallucination outperforms diffusion on protein refolding accuracy, particularly among larger proteins
- design/backbones
- evidence/design-validation
Apr 21, 2026
Including structure prediction confidence while training inverse folding improves sequence diversity but not sequence recovery
Apr 21, 2026
Inverse folding sequence perplexities correlate with Rosetta energies, forward folding TM-scores, and sequence recovery
Apr 21, 2026
Logistic regression outperforms fine-tuned LMs on finding point mutations from NGS data
- design/directed-evolution
- evidence/design-validation
Apr 21, 2026
Low-pLDDT hallucinated models from AF3-generation protein structure predictors are designable
Apr 21, 2026
Monomeric AF2 is better able to distinguish positive and negative PPIs than AF2-multimer
Apr 21, 2026
Most ML quality metrics cannot effectively predict enzyme activity after controlling for similarity to native
Apr 21, 2026
Natural sequences have higher pTM but lower pLDDT than de novo sequences
- evidence/design-validation
- prediction/confidence
Apr 21, 2026
Nearly half of CATH domains do not pass designability filters used to evaluate protein backbone design performance
- evidence/design-validation
- design/backbones
Apr 21, 2026
No one-size-fits-all best approach to motif scaffolding protein design
- design/backbones
- evidence/design-validation
Apr 21, 2026
PLM-designed sequences match the distribution of fitness values, lengths, and structure prediction confidence of natural sequences
- design/sequence-generation
- evidence/design-validation
Apr 21, 2026
PLMs generate more divergent functional proteins than Potts models
- design/sequence-generation
- evidence/design-validation
Apr 21, 2026
Partial diffusion of de novo designed binders and VHHs improves refoldability and designability scores
Apr 21, 2026
Partial structure diffusion can make de novo backbones more designable
- design/backbones
- evidence/design-validation
Apr 21, 2026
Protein structure prediction and design metrics don't correlate with expression probability
Apr 21, 2026
Protein structure prediction methods predict idealized secondary structures in de novo proteins
- prediction/structure
- evidence/design-validation
Apr 21, 2026
Randomly masking residues prior to running ESMFold allows false positives to be identified more effectively
Apr 21, 2026
Residue conservation and solvent exposure data perform comparably to PLMs at some property prediction tasks
Apr 21, 2026
Self-consistency perplexity is correlated with pLDDT
- evidence/design-validation
- prediction/confidence
Apr 21, 2026
Sequence recovery in inverse folding models is not correlated with self-consistency of generated designs
- evidence/design-validation
- design/inverse-folding
Apr 21, 2026
The activity profiles of ancestrally reconstructed enzymes are more difficult to predict than extant or ML-generated enzymes
Apr 21, 2026
The structures of shorter de novo designed sequences are easier to predict than those of longer proteins
- prediction/structure
- evidence/design-validation
Apr 21, 2026
Accurate fitness landscapes are unnecessary for productively engineering enzymes or proteins
- design/directed-evolution
- evidence/design-validation
Apr 21, 2026
Computational models are less designable than experimental structures
- evidence/design-validation
- design/inverse-folding
Apr 21, 2026
High-accuracy computational models might not always be effective for ligand docking
- prediction/ligand-docking
- evidence/design-validation
Apr 21, 2026
High-pLDDT designs can be insoluble
Apr 21, 2026
Inverse folding methods outperform sequence design by hallucination
- design/inverse-folding
- evidence/design-validation
Apr 21, 2026
MD simulations can distinguish enzyme design candidates for catalytically competent substates
Apr 21, 2026
ML-based protein design methods outperform physics-based methods at identifying deleterious sequences but not top sequences
- evidence/design-validation
- design/inverse-folding
Apr 21, 2026
No one-size-fits-all best approach to zero-shot or few-shot protein fitness prediction
- prediction/variant-effects
- evidence/design-validation
Apr 21, 2026
Sequence- and structure-derived ML quality metrics from ML do not correlate with each other
- evidence/design-validation
- prediction/confidence
Nov 04, 2024
Fréchet distance
- evidence/measurements
- evidence/design-validation

Created with Quartz v4.5.2 © 2026

GitHub
Discord Community