archive
Every paper Pith has read. Search by title, abstract, or pith.
48 papers in q-bio.BM · page 1
-
Ensembits turns protein motion into discrete tokens
ENSEMBITS: an alphabet of protein conformational ensembles
-
Hessian matching outperforms force matching on eight proteins
Hessian Matching for Machine-Learned Coarse-Grained Molecular Dynamics
-
Lab variance sets PROTAC prediction ceiling at 0.67 AUROC
Decomposing the Generalization Gap in PROTAC Activity Prediction: Variance Attribution and the Inter-Laboratory Ceiling
-
Deep learning with protein hierarchy predicts complexes better
Deep Learning for Protein Complex Prediction and Design
-
Compact tokenizer generates plausible proteins from scratch with 10x fewer parameters
Yeti: A compact protein structure tokenizer for reconstruction and multi-modal generation
-
New diffusion model generates directional agonists and antagonists
TD3B: Transition-Directed Discrete Diffusion for Allosteric Binder Generation
-
BP180 homotrimer model stays folded in simulations
A putative, computationally stable structure of homotrimeric BP180/collagen XVII
-
Expert fusion model lifts microbial operon accuracy
MicroFuse: Protein-to-Genome Expert Fusion for Microbial Operon Reasoning
-
Graph pooling extracts functional protein substructures from ESM-2 features
Structural Interpretations of Protein Language Model Representations via Differentiable Graph Partitioning
-
SGRPO widens biomolecular utility-diversity frontier
Pushing Biomolecular Utility-Diversity Frontiers with Supergroup Relative Policy Optimization
-
R package MeTime stores all metabolomics steps in one container
MeTime: An R package for reproducible longitudinal metabolomics data analysis
-
Models trained on real folds design RNA under mixed constraints
GoForth: Language Models for RNA Design under Structure, Sequence, and Coding Constraints
-
Masked-position latent prediction beats MLM on protein tasks
ProteinJEPA: Latent prediction complements protein language models
-
Benchmark trains models on noisy DEL screens and tests them on real Ki affinities
CA-DEL: An Open Multi-Target, Multi-Modal Benchmark for Learning from DNA-Encoded Library Screens
-
Tree search guides diffusion to balance multiple protein design goals
MP2D: Constrained Monte Carlo Tree-Guided Diffusion for Multi-Objective Protein Sequence Design
-
AlphaFold guidance raises TM-scores in Phenix cryo-EM models
Enhancing Cryo-EM Density Map Segmentation in Phenix for Improved Atomic Model Building
-
Condensin reverses DNA loops via catch-bond intermediate
Loop Extrusion Reversal by Condensin Motor is Mediated by Catch Bonds
-
Boltz-2 and fine-tuned DrugFormDTA top antiviral affinity benchmarks
Benchmarking open-source tools for in silico antiviral drug discovery
-
Agentic AI matches Smina at 50% docking pose accuracy
AgenticPosesRanker: An Agentic AI Framework for Physically Grounded Ranking of Protein-Ligand Docking Poses
-
Fingerprint injection and anchored chains let small model outperform larger rivals on mole
Bolek: A Multimodal Language Model for Molecular Reasoning
-
ML Re-ranking Boosts Docking Enrichment 110% on LIT-PCBA
Benchmarking Single-Pose Docking, Consensus Rescoring, and Supervised ML on the LIT-PCBA Library: A Critical Evaluation of DiffDock, AutoDock-GPU, GNINA, and DiffDock-NMDN
-
Salt changes BSA scattering via ions and hydration water
Complex Effects of Salt on Small-Angle X-ray Scattering of BSA Originate From the Interplay of Ions and Hydration Water
-
Bayesian method infers nucleic acid motif rates from ligation data
Bayesian Rate Inference for Sequence Motif Dynamics in Systems of Reactive Nucleic Acids
-
Three strategies organize AI work on protein dynamics
Learning Structure, Energy, and Dynamics: A Survey of Artificial Intelligence for Protein Dynamics
-
AI agents train PPI models to 87% accuracy and write matching rules
Agentic AI platforms for autonomous training and rule induction of human-human and virus-human protein-protein interactions
-
Grounding document steers AI coding to valid scientific results
Agentic AI-assisted coding offers a unique opportunity to instill epistemic grounding during software development
-
BioMiner extracts 82k bioactivity entries from 11k papers
BioMiner: A Multi-modal System for Automated Mining of Protein-Ligand Bioactivity Data from Literature
-
Continuous server benchmarks protein function predictors
LAFA: A Framework for Reproducible Longitudinal Assessment of Protein Function Annotation Models
-
Web server detects row and hot mutations missed by standard tools
VARIANT: Web Server for Decoding and Analyzing Viral Mutations at Genome and Protein Levels
-
Channel-wise transforms give AlphaFold3 control over protein conformations
ConforNets: Latents-Based Conformational Control in OpenFold3
-
LLMs meet more molecule constraints when prompts add rules
How Creative Are Large Language Models in Generating Molecules?
-
Parallel MCMC makes protein coupling estimates reproducible
Boltzmann Machine Learning with a Parallel, Persistent Markov chain Monte Carlo method for Estimating Evolutionary Fields and Couplings from a Protein Multiple Sequence Alignment
-
LLM search lifts protein design success rate by 2.5 times
RosettaSearch: Multi-Objective Inference-Time Search for Protein Sequence Design
-
VAEs cluster DNA reactions into pathways and GANs generate cryo-EM maps
Applications of deep generative models to DNA reaction kinetics and to cryogenic electron microscopy
-
Ammonia unlocks nucleobases and fatty acids in Titan crater at 1% level
Prebiotic Chemistry Insights for Dragonfly II: Thermodynamic Favorability of Nucleobases, Ribose, and Fatty Acids in Selk Crater on Titan
-
Chromatin loops follow universal size law set by barriers
Universal Loop Statistics from Active Extrusion with Kinetic Barriers
-
Graph model partitions proteins into functional units
PUFFIN: Protein Unit Discovery with Functional Supervision
-
Distance term doubles recovery of linkable fragment pose pairs
Simultaneous Fragment Docking for Geometrically Linkable Pose Pairs
-
Generative model samples protein shapes at any temperature
Polyformer: a generative framework for thermodynamic modeling of polymeric molecules
-
Smooth scalarization recovers non-convex Pareto fronts in offline RL
Pareto-Optimal Offline Reinforcement Learning via Smooth Tchebysheff Scalarization
-
Protein dynamics lifts mutation effect prediction
TriFit: Trimodal Fusion with Protein Dynamics for Mutation Fitness Prediction
-
Physical forces added to quantum docking QUBO raise accuracy
A Physically-Informed Subgraph Isomorphism Approach to Molecular Docking Using Quantum Annealers
-
Antibody oracles gain 12-20% on developability assays
Biologically-Grounded Multi-Encoder Architectures as Developability Oracles for Antibody Design
-
Dense platelet plugs speed edge fibrin but block core gelation
Platelet plug microstructure and flow modulate fibrin gelation dynamics: Insights from computational simulations
-
Context conditioning enables accurate predictions on targets with only 67 examples
When Does Context Help? A Systematic Study of Target-Conditional Molecular Property Prediction
-
Policy-driven model matches folding rates for 73 proteins
Towards protein folding pathways by reconstructing protein residue networks with a policy-driven model
-
Fine-tuned OCR model matches top sequence accuracy on molecular structures
Fine-tuning DeepSeek-OCR-2 for Molecular Structure Recognition
-
Dual-modal AI identifies 33 shared host factors in flu viruses
ViraHinter: a dual-modal artificial intelligence framework for predicting virus-host interactions