ProteinGym
PulseAugur coverage of ProteinGym — every cluster mentioning ProteinGym across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
ProtModernBERT cuts protein model training costs by 90%
Researchers have developed ProtModernBERT, a new protein language model that significantly reduces training costs while maintaining competitive performance. This modernized BERT-style model is designed for resource-cons…
-
AI advances protein design with language guidance and security focus · 6 sources tracked
Researchers are exploring new methods for protein design using advanced AI models, focusing on natural language guidance and data integrity. One study details two data-only poisoning attacks that can manipulate protein …
-
New benchmark quantifies biosecurity risk of LLM-driven biological mutation design
Researchers have developed DRYLAB-Bench, a new benchmark designed to evaluate the biosecurity risks associated with large language models (LLMs) capable of biological mutation design. This benchmark assesses LLM agents …
-
New hypothesis and search method improve protein evolution with language models
Researchers have introduced a new hypothesis called the Linear Fitness Subspace (LFS), which suggests that a small set of directions within protein language models (PLMs) can accurately predict fitness variations for pr…
-
StabilityArc predicts protein stability landscapes with cross-protein transfer
Researchers have developed StabilityArc, a novel method for predicting protein stability landscapes. This approach uses a shared decoder to interpret biochemical constraints across different proteins, enabling better pr…
-
New Q-BIOLAT framework optimizes protein fitness landscapes using binary codes
Researchers have developed Q-BioLat, a new framework for optimizing protein fitness landscapes. This method maps protein language model embeddings to compact binary codes, enabling the use of quadratic unconstrained bin…
-
Tabular foundation models show surprising generalization to biomolecular prediction tasks
A new research paper explores the surprising effectiveness of tabular foundation models, such as TabPFN and TabICL, in predicting biomolecular properties. Despite being pretrained on synthetic data with no direct link t…
-
New DPLM-Evo model enables generative protein evolution with explicit edit predictions
Researchers have developed DPLM-Evo, a novel evolutionary discrete diffusion framework designed to better model protein evolution. Unlike previous models that used masking, DPLM-Evo explicitly predicts substitution, ins…