Community benchmark tests 12,000 AI-designed mini-proteins in living CAR T cells
A global competition named Bits to Binders evaluated 12,000 AI-designed 80-amino acid mini-proteins as the targeting heads for CAR T-cell therapies against human CD20.
Community benchmark tests 12,000 AI-designed mini-proteins in living CAR T cells
A global competition named Bits to Binders evaluated 12,000 AI-designed 80-amino acid mini-proteins as the targeting heads for CAR T-cell therapies against human CD20.
Machine learning guides directed evolution to speed up plant Rubisco
Researchers used the structure-informed protein language model ESM-IF1 to pick mutation sites in tobacco Rubisco, the primary enzyme plants use to fix carbon dioxide.
Combining structural models with sequence AI improves protein fitness predictions under distribution shifts
A machine-learning framework augments protein language model representations with structural priors and evolutionary co-variation features derived from an AlphaFold-style trunk.
Machine learning boosts directed evolution yields for engineered viral capsids 41-fold
Researchers combined machine learning with directed evolution to engineer adeno-associated virus (AAV) capsids optimized for exosomal encapsulation, a method used to protect gene therapy delivery vehicles from immune recognition.
Small shifts in AI protein embeddings let algorithms find fitter designs with fewer tests
Researchers showed that assay-specific fitness variations align within a compact, linear subspace of mutation representation changes produced by protein language models.
Generative model designs proteins, mRNA, and DNA simultaneously
Researchers built NuCaliby, a structure-conditioned generative model that designs protein amino-acid sequences together with their corresponding DNA and mRNA nucleotide sequences.
Protein language model predicts how molecular motors turn fuel into mechanical work
Researchers developed CatFunc, a lightweight protein language model pipeline that classifies how AAA proteins convert ATP chemical energy into mechanical force directly from sequence data.
Generative protein models help design targeted enzymes that strip phosphate tags from cancer and tau proteins
Researchers have developed peptide-retargeted phosphatases, synthetic fusion proteins that link computationally designed peptide binders to a serine/threonine phosphatase subunit to selectively remove phosphate groups from specific cellular targets.
Diffusion guidance helps build functional enzymes with multi-residue active sites
Bioengineers developed the Catalytic Guidance Engine, an iterative design framework that detunes geometric constraints during diffusion sampling to build complex catalytic sites into de novo proteins.
Latent vocabularies help autoregressive models generate proteins faster
Researchers introduced two learned discrete latent vocabularies to train autoregressive transformers on protein sequence and structure, replacing conventional residue tokens and coordinates.
AI pipeline designs working enzymes from scratch for new-to-nature chemical reactions
Researchers introduced AlphaProtein Novo, a machine-learning pipeline that builds functional enzymes from scratch without starting from natural protein backbones.
Machine learning and simulations discover natural enzymes for green plastics
Researchers combined molecular simulations and machine learning across four screening rounds to prospect natural hydroxymethylfurfural oxidases for synthesizing the bioplastic precursor 2,5-furandicarboxylic acid (FDCA).
Diffusion model sketches the shifting shapes of disordered proteins
A geometric diffusion model called Zephyr samples conformational ensembles of intrinsically disordered proteins directly from sequence, bypassing the need for slow physics simulations.
Bayesian method accounts for experimental noise in protein fitness training data
Researchers built FLIGHTED, a Bayesian statistical method that converts noisy high-throughput assay readouts into probabilistic fitness landscapes before feeding them into machine-learning models.
Generative model links protein sequence and 3D structure to track evolution
Researchers combined amino acid alignments with a structural alphabet, using variational autoencoders and evolutionary energy calculations to map protein sequence variations against structural changes.
Xaira introduces its antibody design platform and first preclinical oncology assets
Xaira Therapeutics unveiled X-Design, its generative artificial intelligence platform built to design drug-ready antibodies from scratch, along with two early preclinical pipeline candidates.
Compressed protein language models retain over 90% of original performance
A systematic benchmark evaluated 4-bit quantization and low-rank adapter fine-tuning, known as QLoRA, across seven protein language models including ESM-2, ProtT5, and Profluent-E1.
AI pipeline builds 3D protein structures straight from microscope images
Fold'EM combines generative structural priors directly with raw cryo-electron microscopy particle images to infer atomic models, skipping the usual steps of building a 3D density map and fitting a model into it.
AlphaFold models unmapped protein pairings across 2,800 pandemic-threat viruses
Researchers ran AlphaFold2 and AlphaFold-Multimer across 2,812 viral proteomes from 23 viral families, generating high-confidence models for 5,279 heterodimers and 2,749 homodimers.
Protein design tool engineers acid-sensitive binders directly from sequence
Proton-PottsMPNN treats protonated and deprotonated amino acids as distinct sequence tokens, allowing researchers to design proteins that bind their targets only at specified acidic pH levels.
Screening tool speeds up protein complex model selection 300-fold
DeepUMQA-Selection pairs single-model accuracy scores with shape-recognition structural alignment to pick top predicted multi-protein structures from large candidate pools.
Biological context helps general AI models match specialized protein tools
Researchers built BioEvidence, an interface that feeds structural and evolutionary context from standard biology tools into frozen large language models to predict how protein mutations affect function.
AlphaFold3 filters broad-spectrum nanobodies against snake venoms
Researchers combined high-throughput yeast display screening with AlphaFold3 structural modeling to discover single-domain antibodies that neutralize neurotoxins from multiple snake species.
Combining sequence and structure models sharpens protein interaction predictions
Researchers built PeTriPPI2, a computational framework that predicts whether two proteins interact by pairing sequence embeddings from ESM-2 with structural features from PeTriBERT.
Watermarking system tags AI-designed proteins without reducing their function
Google researchers introduced SynthIDBio, a set of methods that embeds detectable digital watermarks directly into AI-generated protein sequences and molecular structures.
Designed mini-proteins block malaria parasite invasion at nanomolar potency
Computational design yielded de novo mini-proteins that target Plasmodium falciparum reticulocyte-binding protein homolog 5 (RH5) to block its interaction with human basigin, an essential step in red blood cell invasion.
Guiding antibody generation models with known binding fingerprints produces high-affinity binders
Researchers introduced MIMOSA, an inference-time framework that constrains generative protein diffusion models to mimic the interaction geometry of known binding partners.
Training-free harness boosts protein foundation model accuracy across 3 million mutations
Researchers introduced VenusREM-Harness, a plug-in method that improves mutation fitness predictions from frozen protein foundation models without retraining.
Simple math fix boosts accuracy across ten protein structure models
Researchers at the Oxford Protein Informatics Group derived a closed-form correction for a mathematical bias caused by pairing Dropout and LayerNorm layers, a common design in AlphaFold2-style architectures.
Graph neural network learns transferable features of protein complexes
Researchers introduced GraPPI, a self-supervised deep learning model designed to represent protein-protein interactions by treating complexes as heterogeneous residual graphs.