Tag
OpenAI introduces GeneBench-Pro, a research-level benchmark designed to test AI agents' ability to perform judgment-heavy analyses in computational biology, covering genomics, quantitative biology, and translational medicine.
This study uses single-cell RNA sequencing to map the developmental trajectory of human adipocytes, identifying 15 cell clusters and key signaling pathways such as IGF and FGF, offering potential therapeutic targets for obesity and metabolic disorders.
This paper proposes a framework linking partially observable Markov decision processes (POMDPs) with biochemical reaction dynamics to model phototaxis in unicellular algae, using inverse reinforcement learning to infer behavioral objectives from experimental trajectories.
Nabla Bio unveils JAM-2, a model for zero-shot drug design achieving atomic-precision, computationally designed multispecific antibodies and dual-variant KRAS multispecifics with high potency and selectivity, validated with Cryo-EM and wet-lab experiments.
Introduces ProHiFlo, a hierarchical flow matching framework for de novo protein generation with coarse-to-fine generation, functional guidance, and SE(3)-equivariant architecture, achieving state-of-the-art performance with 4x fewer sampling steps.
Anthropic's science blog argues that AI progress in biology lags behind coding because biological data infrastructure is not designed for agents. A case study shows that adding a deterministic retrieval layer (gget virus) boosts accuracy to nearly 100%.
Introduces ShallowBench, a curated benchmark of 5,780 shallow-pocket protein targets, to evaluate generative drug design models on challenging low-pocketability targets like KRAS and MYC.
BlockFormer introduces a transformer architecture for solving inverse problems from block-structured interaction maps, such as centromere identification from Hi-C data, using a custom simulator for synthetic training data.
This paper formalizes transcriptome-based drug design (TBDD) as a generative inverse problem and proposes CURE, a multi-resolution transcriptome-guided diffusion framework that generates drug molecules conditioned on desired transcriptomic state transitions.
This PhD thesis introduces deep learning methods for protein complex prediction and design, including GLINTER for contact prediction, ESMPair for homolog pairing, and RedNet for binder design.
This paper introduces a new paradigm for universal Gene Regulatory Network (GRN) inference using single-cell foundation models, proposing Virtual Value Perturbation and Gradient Trajectory methods to distill regulatory knowledge.
This article serves as a beginner's guide to protein lead optimization in drug design, specifically explaining the Cradle-1 pipeline and foundational concepts of protein structure and function.
This paper introduces a discrete diffusion model with a novel 'germline absorbing' modification to improve conditional antibody sequence generation. It addresses germline bias in protein language models and demonstrates superior performance in optimizing antibody binding affinity and developability compared to existing methods like EvoProtGrad.
This article details the development of CodonRoBERTa, a language model trained across 25 species for mRNA codon optimization, highlighting a cost-effective pipeline that includes protein folding and sequence design.
MIT researchers have developed VibeGen, an AI model that designs proteins based on their dynamic motion and mechanics rather than just static structure. This approach allows for the creation of proteins with specific vibrational and flexing behaviors, advancing the field of generative AI in science.
AI tools like ChatGPT are shortening antibiotic discovery from years to hours through computational molecular design, addressing the global antimicrobial resistance crisis.