· Ravit Netzer
Predicting Thermostability Directly from Protein Sequence
How Scala's stability model moves from amino acid sequence to a predicted melting temperature shift without requiring solved structure.
Read articleWriting from the Scala Biodesign team on stability prediction, machine learning for proteins, and practical computational biology.
Featured Article
· Ravit Netzer
How Scala's stability model moves from amino acid sequence to a predicted melting temperature shift without requiring solved structure.
Read article
· Oren Katz
Why scanning the full combinatorial variant space first saves weeks of back-and-forth at the synthesis stage.
Read article
· Tamar Levy
A look at how prediction accuracy in our early-access program shifted as we incorporated feedback from more diverse protein families.
Read article
· Ravit Netzer
Breaking down where computational pre-screening saves the most time and where wet-lab validation is still non-negotiable.
Read article
· Oren Katz
The tension between solubility and thermal stability is one of the trickiest engineering tradeoffs. Here is how our model handles it.
Read article
· Tamar Levy
Single-point mutation predictions miss epistatic effects. We describe how Scala accounts for pairwise and higher-order interactions in variant scoring.
Read article
· Oren Katz
A technical comparison of physics-based energy function approaches and where each one gains or loses accuracy on real stability benchmarks.
Read article
· Ravit Netzer
Results from one of our first extended pilot runs: a lipase variant panel where the top-ranked computational candidates matched the wet-lab outcomes.
Read article
· Tamar Levy
An overview of how fitness landscape representations help protein engineers navigate large variant spaces toward higher-stability solutions.
Read article
· Oren Katz
The architectural decisions behind Scala's core pipeline: from raw sequence input through feature extraction to a stability delta score output.
Read article
· Ravit Netzer
The mutation scanner lets you map stability predictions across your full single-point and combinatorial variant space in a single run.
Read article
· Tamar Levy
Why we built Scala's models on a hybrid foundation of physics-informed energy functions and sequence-fitness learned features, and why that matters for accuracy.
Read article
· Ravit Netzer
Stability failures are the leading cause of attrition in early enzyme and antibody engineering campaigns. Here is how computational tools change the economics.
Read article
· Oren Katz
Reflecting on the product decisions that shaped Scala's interface: why we optimized for wet-lab scientists who are not machine learning experts.
Read article