https://hstgenomics.com/news.html AI developments in bioinformatics, drug discovery and the life sciences, linked to the original. en-gb https://arxiv.org/abs/2610.00742hst-news-1354Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTEvery protein has a unique stability landscape, but the physical consequences of mutation are governed by recurring biochemical constraints. We test whether a shared decoder, trained on measurements from diverse proteins, can interpret these constraints in an unseen target, enabling cross-protein transfer for initial experimental round prescreening. We present StabilityArc , which maps frozen ESMC-600M residue representations through a shared RoPE transformer to an Lx20 matrix of substitution effects; a symmetric, contact-aware residual aids in predicting epistasis in simultaneous substitution | Aaron L. Feller, Andrew D. Ellington, Claus O. Wilke | arXiv · 2026-10-02 https://arxiv.org/abs/2610.01086hst-news-1356Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTPeptides occupy a particularly challenging regime of biomolecular dynamics. Short peptides lack a stable folded core and populate broad conformational ensembles, while cyclization adds ring closure, non-local residue coupling, and stereochemical diversity. Deep generative models have made remarkable progress in emulating molecular-dynamics trajectories directly, yet each is trained on windows cut at a fixed interval and therefore observes the process at a fixed temporal resolution. More fundamentally, none takes that interval as an input, so the physical time a window spans is never represente | Xichen Sun, Wentao Wei, Xiaoxi Zhang et al. (5 authors) | arXiv · 2026-10-02 https://arxiv.org/abs/2610.01358hst-news-1357Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTSingle-particle cryo-electron microscopy (cryo-EM) has become a widely adopted technique for biomolecular structure determination. The conventional cryo-EM computational pipeline first combines many particle images to reconstruct an electrostatic potential (ESP) map and then fits an atomic model to the recovered map. Density reconstruction has high sample complexity, requiring large numbers of particle images and making structure determination high-cost and low-throughput, particularly for heterogeneous samples. Downstream atomic model building, in turn, becomes increasingly difficult as the r | Advaith Maddipatla, Märt-Erik Mäeots, Marco Pegoraro et al. (8 authors) | arXiv · 2026-10-02 https://arxiv.org/abs/2610.00191hst-news-1358Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTModeling protein-protein interactions requires accurate scoring functions that can rank potential poses (conformations) of a protein-protein complex to differentiate near-native poses from incorrect ones. Here, we propose a general framework for improving protein-protein pose ranking and other biomolecular interaction models using the LambdaLoss loss function from the Learning-to-Rank field. We test this framework by fine-tuning the energy prediction head of DFMDock with the LambdaLoss on an augmented dataset of 2.9M decoy poses derived from the DIPS dataset. On targets from the CAPRI score se | Richard Zhu, Darren Xu, Lee-Shin Chu et al. (4 authors) | arXiv · 2026-10-02 https://arxiv.org/abs/2610.00988hst-news-1359Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTCryptic ligand-binding pockets are not apparent in experimentally determined apo structures, making them difficult to identify from unbound receptor geometry. A complementary challenge is to make the structural measurements and learned evidence behind each prediction directly inspectable. We introduce a supervised algebraic counting field (ACF) for predicting cryptic-pocket residues from apo structures. ACF compiles explicit geometric, physicochemical, and topological features into compact, integer-weighted lookup tables. Each prediction score can be reconstructed from feature values, training | Shan Yu, Xuening Wu | arXiv · 2026-10-02 https://arxiv.org/abs/2610.01663hst-news-1360Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTBiomolecular therapeutics often start from known sequences and require targeted editing to improve multiple properties while satisfying hard biochemical and manufacturability constraints. However, existing generative methods do not jointly support multi-objective optimization, hard feasibility, and sequence editing in discrete, variable-length biological spaces. In this work, we introduce Pareto-Constrained Molecule Editing (pCoMole), a framework built on discrete flow matching that steers a pre-trained Edit Flow toward user-specified preferences while enforcing terminal feasibility. pCoMole d | Tong Chen, Maximilian Holsman, Lin Zhao et al. (4 authors) | arXiv · 2026-10-02 https://arxiv.org/abs/2610.01891hst-news-1361Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTProtein mutation effect generation asks a model to describe the functional consequence of a point mutation in natural language. Existing protein-to-text systems typically encode mutation information into undifferentiated representations, overlooking the organization of mutation-induced evidence across structural and biochemical factors. We propose RipplePLM, a mutation-aware generation framework centered on Direct-Distal Cross-Attention (DDCA). By constructing a residue-level Mutation Perturbation Field from pre-trained protein language models, DDCA leverages predicted contact maps to organize | Liuzhenghao Lv, Yuyang Liu, Yuyang Gao et al. (5 authors) | arXiv · 2026-10-02 https://arxiv.org/abs/2610.00283hst-news-1362Genomics & single-cellThu, 01 Oct 2026 22:00:00 GMTThe extensive diversity of viral genomes challenges alignment-free representations to capture both nucleotide composition and the higher-order organization of recurring sequence patterns. Conventional $k$-mer representations quantify pattern abundance but do not explicitly encode how these patterns are organized across scales. Here, we introduce CANN, an alignment-free framework that combines persistent commutative-algebraic descriptors with neural representation learning for viral family classification. CANN uses persistent facet descriptors to encode the multiscale organization of $k$-mer pa | Mushal Zia, Faisal Suwayyid, Guo-Wei Wei | arXiv · 2026-10-02 https://www.biorxiv.org/content/10.64898/2026.09.27.754743v1hst-news-1422Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTStructure-guided antibody CDR design constrains residue selection using backbone geometry, but structural evidence alone may leave multiple amino acids plausible. We investigate whether the fixed antibody framework provides complementary sequence information for resolving this ambiguity. We introduce AbRefine, an information-aware fusion framework that combines backbone-conditioned structural predictions with framework-conditioned sequence distributions derived from a pretrained antibody language model with all designed CDRs masked. Its Information-Aware Fusion (IAF) module learns residue-spec | Qi Liu, Xinyi Xu, Qian Cheng et al. (8 authors) | bioRxiv · 2026-10-02 · DOI 10.64898/2026.09.27.754743 https://www.biorxiv.org/content/10.64898/2026.09.26.754739v1hst-news-1423Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTProtein complex structure prediction depends on both the breadth of candidate coverage and the ability to select accurate models within a limited output budget. Most existing approaches emphasize only one stage. End-to-end models and molecular docking workflows generate candidate structures, whereas quality-assessment models primarily rerank a predefined candidate pool. A complete strategy must account for complementary sampling across generators, differences among scoring scales and the allocation of final candidate quotas. Here, EnsPlex is presented as a multi-source framework that couples s | Baochuan Hu, Zhenhua Lu, Khalid Zaman et al. (5 authors) | bioRxiv · 2026-10-02 · DOI 10.64898/2026.09.26.754739 https://www.biorxiv.org/content/10.64898/2026.09.27.754741v1hst-news-1424Genomics & single-cellThu, 01 Oct 2026 22:00:00 GMTMotivation: LLM agents increasingly draft functional interpretations of gene lists, but can cite Gene Ontology (GO) terms that no current enrichment backend returned for that list, and can pair real GO accessions with fabricated labels. Results: We present GroundAnnot, a client for PANTHER, Enrichr, and g:Profiler that returns a closed vocabulary of GO term IDs and their backend labels, and enforces two contracts: Contract A (no ID or label outside the backend payload) and Contract B (no enrichment claim outside the backend's significant set). Across three open-weight models (Qwen2.5-7B local; | Maano Bryton Malima | bioRxiv · 2026-10-02 · DOI 10.64898/2026.09.27.754741 https://www.biorxiv.org/content/10.64898/2026.09.27.754054v1hst-news-1425Genomics & single-cellThu, 01 Oct 2026 22:00:00 GMTTranscriptomic classification is often hindered by the small number of samples relative to the high dimensionality of gene expression data. We introduce General Gene Expression (GGE), a deep meta-learning framework designed to support robust classification in this limited-sample setting. By training across 5,220 distinct biomedical prediction objectives drawn from 1,779 different human datasets, GGE learns a model initialization that captures biological patterns shared across heterogeneous classification tasks. This learned initialization has two key advantages. First, it enables improved perf | Gal Yankovitz, Irit Gat-Viks | bioRxiv · 2026-10-02 · DOI 10.64898/2026.09.27.754054 https://www.biorxiv.org/content/10.64898/2026.09.27.754758v1hst-news-1426Proteins & molecular designThu, 01 Oct 2026 22:00:00 GMTProtein language models have become a dominant foundation for sequence-based protein-protein interaction (PPI) prediction, but their generalization remains limited under stringent evaluation, and benchmark accuracy alone cannot reveal whether a PPI predictor learns partner-specific biological signals or exploits contextual shortcuts. Here we introduce AuditPPI, an interpretable framework that transforms sparse-autoencoder (SAE) features from a frozen protein language model into order-invariant pair fingerprints, recasting PPI prediction as an auditable tabular-learning problem. AuditPPI achiev | Weimin Zhu, Shengfan Wang, Xiaojian Liu et al. (7 authors) | bioRxiv · 2026-10-02 · DOI 10.64898/2026.09.27.754758 https://www.nature.com/articles/s41587-026-03307-whst-news-1364Genomics & single-cellWed, 30 Sep 2026 22:00:00 GMTCalibration-aware evaluation of perturbation models shows that deep learning models can outperform baselines. | Henry E. Miller, Gabriel M. Mejia, Francis J. A. Leblanc et al. (6 authors) | Nature Biotechnology · 2026-10-01 · DOI 10.1038/s41587-026-03307-w https://www.nature.com/articles/s41592-026-03243-2hst-news-1372Drug discoveryWed, 30 Sep 2026 22:00:00 GMTA method for transferable retention time prediction across chromatographic conditions and compound classes is presented. The approach makes use of a two-step strategy that first predicts a condition-aware retention order index followed by mapping it to absolute retention times. | Fleming Kretschmer, Eva-Maria Harrieder, Michael Witting et al. (4 authors) | Nature Methods · 2026-10-01 · DOI 10.1038/s41592-026-03243-2 https://www.biorxiv.org/content/10.64898/2026.09.30.755756v1hst-news-1403Genomics & single-cellWed, 30 Sep 2026 22:00:00 GMTVirtual cell models aim to predict how cells respond to perturbations, yet accurate prediction of post-perturbation gene expression can conceal failures to recover context-specific responses. Here we identify context geometry compression (CGC), in which models preserve cellular identity and shared perturbation effects while collapsing differences in how the same perturbation acts across contexts. Across biological systems, reproducible context-specific responses remained difficult to recover; in a large drug-response matrix, this persisted even with nearly complete perturbation coverage. The s | Yongqi Huang, Hanzhi Wang, Chenyu Li et al. (4 authors) | bioRxiv · 2026-10-01 · DOI 10.64898/2026.09.30.755756 https://www.biorxiv.org/content/10.64898/2026.09.27.754822v1hst-news-1406Proteins & molecular designWed, 30 Sep 2026 22:00:00 GMTLarge language models (LLMs) show emerging zero-shot capability for protein variant prediction, yet still lag behind specialized protein models. We ask whether this gap can be reduced by scaling access to biological evidence rather than adapting model parameters. We introduce BioEvidence, a training-free and model-agnostic interface that converts structural and evolutionary information from standard biological tools into compact evidence for frozen LLMs. On the ProteinGym benchmark, we observe evidence scaling: performance improves as evidence becomes richer. Structural and evolutionary eviden | Zitong Hao, Chaoyang Wang, Dongyuan Li et al. (4 authors) | bioRxiv · 2026-10-01 · DOI 10.64898/2026.09.27.754822 https://www.biorxiv.org/content/10.64898/2026.09.25.754484v1hst-news-1409Genomics & single-cellWed, 30 Sep 2026 22:00:00 GMTSpatial gene expression prediction from histology is typically evaluated by the mean per-gene Pearson correlation (PCC). PCC does not distinguish spatial scales: coarse tissue organization alone can earn most of the score. A domain oracle makes this precise: given an image-derived partition of the tissue, the predictor that assigns each domain its measured mean expression is the best that any predictor constant on that partition can do. On a widely used benchmark with a ~100 um pitch, a learnable version that estimates each domain mean from training data reaches 82% of the PCC of a trained mod | Qing Wang, Sipei Gu, Qizhen Lan et al. (6 authors) | bioRxiv · 2026-10-01 · DOI 10.64898/2026.09.25.754484 https://www.biorxiv.org/content/10.64898/2026.09.25.754557v1hst-news-1411Proteins & molecular designWed, 30 Sep 2026 22:00:00 GMTAntidiabetic peptides (ADPs) are short bioactive sequences of therapeutic interest, and sequence-based classifiers prioritise experimental candidates. A classifier is useful only if its accuracy transfers to unseen sequences, which depends on how the benchmark was as sembled and partitioned. We term the distance between what such a classifier is scored on and the function it is meant to predict the provenance-to-function gap, and we quantify it. We re-evaluate the two-layer ADP benchmark of Basith et al. under a protocol that groups homologous peptides rather than splitting them at random. Of | Ariful Islam, Md. Faruk Hosen, Md. Abul Basar et al. (5 authors) | bioRxiv · 2026-10-01 · DOI 10.64898/2026.09.25.754557 https://www.biorxiv.org/content/10.64898/2026.09.26.754618v1hst-news-1412Proteins & molecular designWed, 30 Sep 2026 22:00:00 GMTIn the post-AlphaFold era, advances in protein complex structure prediction have enabled the generation of thousands of candidate models per target, making the accurate and efficient selection of high-quality models from large candidate pools a critical challenge. Existing model-selection approaches commonly employ estimated model accuracy (EMA) methods, which include single-model and consensus-based methods. The former lack cross-model comparative information, whereas the latter often incur high computational costs from pairwise structural alignments and are influenced by candidate-pool quali | Fang Liang, Lei Xie, Enjia Ye et al. (9 authors) | bioRxiv · 2026-10-01 · DOI 10.64898/2026.09.26.754618 https://www.biorxiv.org/content/10.64898/2026.09.29.755321v1hst-news-1418Proteins & molecular designWed, 30 Sep 2026 22:00:00 GMTMechanistic understanding of viral protein-protein interactions enables global health security and pandemic preparedness, yet experimental characterisation remains difficult, costly, and often restricted to specialised laboratories. Here, we describe an in silico campaign using AlphaFold2 and AlphaFold-Multimer to predict monomers and dimer structures within 2,812 viral proteomes from 23 viral families relevant to human health. We report high-confidence predicted structures of 5,279 hetero-, and 2,749 homo-dimers. Structural clustering of interfaces reduces the set fivefold to 1,598 consolidat | Yewon Han, Riya Narain, Razan Abbara et al. (35 authors) | bioRxiv · 2026-10-01 · DOI 10.64898/2026.09.29.755321 https://blog.google/innovation-and-ai/models-and-research/google-research/google-science-ai-flu-forecasts/hst-news-1268Medicine, imaging & neuroscienceWed, 30 Sep 2026 22:00:00 GMTGoogle’s science AI model was the best at forecasting flu-related hospital admissions, the Centers for Disease Control announced. | Google · 2026-10-01 https://blog.google/innovation-and-ai/models-and-research/google-deepmind/synthid-bio/hst-news-1270Biosecurity & industryTue, 29 Sep 2026 22:00:00 GMTGoogle DeepMind introduces SynthID Bio to watermark AI-designed proteins while maintaining biological function. | Google · 2026-09-30 https://www.nature.com/articles/s41592-026-03224-5hst-news-1373Genomics & single-cellTue, 29 Sep 2026 22:00:00 GMTCytoVI is a deep generative model for statistically rigorous analysis of antibody-based single-cell data. | Florian Ingelfinger, Nathan Levy, Can Ergen et al. (15 authors) | Nature Methods · 2026-09-30 · DOI 10.1038/s41592-026-03224-5 https://www.biorxiv.org/content/10.64898/2026.09.17.752359v1hst-news-1382Genomics & single-cellTue, 29 Sep 2026 22:00:00 GMTAnticipating how mutations change viral fitness is central to genomic surveillance and vaccine design, yet the supervised phenotype data behind the most accurate variant-effect predictors are unavailable for most emerging pathogens. We ask how far label-free scoring can go using only sequences, their evolutionary history, and structure. We assemble a modular, fully unsupervised pipeline that estimates a few interpretable terms (intrinsic replicative fitness, antigenic escape, and realized growth), and that lets each term be produced by more than one estimator, so the estimator itself becomes a | Stephane Aris-Brosou, Matthieu Vilain | bioRxiv · 2026-09-30 · DOI 10.64898/2026.09.17.752359 https://www.biorxiv.org/content/10.64898/2026.09.25.754372v1hst-news-1386Genomics & single-cellTue, 29 Sep 2026 22:00:00 GMTSpatial transcriptomics reveals cellular organization within intact tissues, whereas dissociated single-cell RNA sequencing provides broad transcriptomic coverage but loses information about cellular proximity and neighborhood structure. Here, we developed ProxiNet, a spatially supervised framework that learns transcriptomic signatures of pairwise cellular proximity from spatial reference datasets and transfers these relationships to dissociated single-cell transcriptomes. ProxiNet predicted cellular proximity across brain regions and spatial technologies, including zero-shot cross-technology | Yuxiang Zhan, Binwei Yan, Ao Zhang et al. (5 authors) | bioRxiv · 2026-09-30 · DOI 10.64898/2026.09.25.754372 https://www.biorxiv.org/content/10.64898/2026.09.29.755338v1hst-news-1388Proteins & molecular designTue, 29 Sep 2026 22:00:00 GMTRecent advances in artificial intelligence have enabled models to capture protein sequence and structural features. We present PeTriPPI2, a hybrid framework for protein--protein interaction (PPI) prediction that combines sequence embeddings from ESM-2 with structural representations from PeTriBERT, an encoder originally designed for inverse folding. We evaluate this sequence--structure approach on the Pinder dataset. On a test set of 2,342 protein pairs, the selected checkpoint achieves 0.898 accuracy, 0.917 precision, 0.876 recall, and 0.896 F1 score for the interacting class (AUROC 0.963). C | Arthur Wavreille, Gabriel Krouk, Baldwin Dumortier | bioRxiv · 2026-09-30 · DOI 10.64898/2026.09.29.755338 https://www.biorxiv.org/content/10.64898/2026.09.24.754260v1hst-news-1389Genomics & single-cellTue, 29 Sep 2026 22:00:00 GMTAccurate cell identity annotation is essential for reconstructing cell lineages and gene-regulatory networks in embryogenesis. However, in Caenorhabditis elegans (C. elegans) single-cell RNA-sequencing (scRNA-seq) atlases, closely related sub-lineage cells often have near-identical transcriptomes, leading to merged labels and limiting downstream analyses. We present CHACAM (Cell-Cell Interaction-guided Hierarchical Attention-based Cell Allocation Model), a supervised machine learning framework that integrates gene expression, a curated C. elegans ligand-receptor (L-R) interaction database, and | Xingyu Chen, Xinxin Ju, Maithri Murali et al. (6 authors) | bioRxiv · 2026-09-30 · DOI 10.64898/2026.09.24.754260 https://www.biorxiv.org/content/10.64898/2026.09.24.754263v1hst-news-1390Genomics & single-cellTue, 29 Sep 2026 22:00:00 GMTSingle-cell RNA sequencing (scRNA-seq) has transformed our understanding of cellular heterogeneity, yet most datasets are limited to a handful of model organisms, leaving critical gaps in cross-species biology. Existing computational methods for integrating multi-species scRNA-seq data often rely on one-to-one ortholog mapping, which fails to account for gene duplications, losses, or functional divergences. To address these challenges, we introduce CycloCross, an adversarial method based on the CycleGAN architecture, designed to translate scRNA-seq data between species without requiring a prio | Olympio Hacquard, Yuya Tokuta, Yusuke Imoto et al. (8 authors) | bioRxiv · 2026-09-30 · DOI 10.64898/2026.09.24.754263 https://huggingface.co/papers/2609.40359hst-news-329Medicine, imaging & neuroscienceTue, 29 Sep 2026 22:00:00 GMTWe find that major reported improvements in decoding words from non-invasive brain recordings are largely reproducible without any brain data. In the influential work of d'Ascoli et al. (2025), time series of brain activity from subjects perceiving continuous speech are segmented into fixed-length windows starting at each word. A neural network then generates predictions for all of the words in a sentence together. Neighbouring windows partially overlap, implicitly revealing the interval between words. Since these intervals indicate the duration of the words spoken, and different words tend to | Dulhan Jayalath, Oiwi Parker Jones | arXiv · 2026-09-30 https://www.microsoft.com/en-us/research/blog/introducing-quine-an-ai-research-system-designed-for-the-complexity-of-biology/hst-news-182AI models & research toolsMon, 28 Sep 2026 22:00:00 GMTBiology doesn't operate in silos, and neither should the AI representation of it. Quine is an early-stage research effort to create a multimodal world model of biology. By connecting insights across biological scales and modalities, Quine helps scientists computationally search a space far larger than intuition allows and prioritize hypotheses before they reach the lab. Experimental results provide important feedback, helping researchers sharpen future research directions. | Microsoft Research · 2026-09-29 https://www.isomorphiclabs.com/articles/building-a-new-path-to-make-medicines-with-aihst-news-1363Drug discoveryMon, 28 Sep 2026 22:00:00 GMTIsomorphic Labs · 2026-09-29 https://www.nature.com/articles/s41587-026-03330-xhst-news-1367Drug discoverySun, 27 Sep 2026 22:00:00 GMTAGENT, an AI-driven framework integrating high-throughput experimentation with Bayesian optimization, enables the rapid engineering of thermostable, solid-state mRNA vaccines. These formulations retained full bioactivity after storage at 37 °C for 2 months, elicited immune responses non-inferior to fresh vaccines, and enabled cold-chain-free delivery via microneedle patches in rodents and nonhuman primates. | Nature Biotechnology · 2026-09-28 · DOI 10.1038/s41587-026-03330-x https://huggingface.co/papers/2609.34274hst-news-313AI models & research toolsSun, 27 Sep 2026 22:00:00 GMTArtificial-intelligence (AI) agents hold promise for automating bioimage analysis, yet no benchmark evaluates whether they can carry out real-world analyses end to end. Such analyses are hard for agents because 2D images, 3D volumes and time-lapse sequences are often too large to read as context, so an agent must choose and run an analysis through code, specialized software and rendered views. Published studies make this capability testable, because each pairs raw images with a peer-reviewed result. We introduce BIABench, a benchmark of 16 tasks reconstructed from published biological studies | Zixuan Pan, Davide Panzeri, Lukas Johanns et al. (8 authors) | arXiv · 2026-09-28 https://huggingface.co/papers/2609.32810hst-news-231Medicine, imaging & neuroscienceFri, 25 Sep 2026 22:00:00 GMTMultidisciplinary tumor boards integrate multimodal clinical observations and longitudinal patient histories through specialist discussions, yet benchmarks rarely capture these real-world trajectories. We introduce OpenTumorBoard, a benchmark with 611 patient cases and 19,157 discussion turns across ten specialist roles, transcribed from 12,534 minutes of publicly available tumor board recordings on YouTube. The benchmark evaluates two settings: SPECIALIST TURN, in which an LLM responds to a clinically significant question posed during a real discussion, and BOARD SIMULATION, in which it gener | Anqi Li, Zhixuan Ge, Yixuan Duan et al. (12 authors) | arXiv · 2026-09-26 https://developer.nvidia.com/blog/efficient-moe-training-for-biological-foundation-models/hst-news-389AI models & research toolsWed, 23 Sep 2026 22:00:00 GMTAs language models grow, scaling dense architectures becomes increasingly expensive. In a dense transformer, every token passes through every layer, so adding... | NVIDIA · 2026-09-24 https://www.nature.com/articles/s41592-026-03212-9hst-news-1376Genomics & single-cellWed, 23 Sep 2026 22:00:00 GMTSegmentation and quantification of spatial transcriptomics data is plagued by specific noise and bias. ResolVI tackles this challenge by generating error- and batch-corrected probabilistic representations of spatial transcriptomics data, contributing to improved performance in multiple analysis tasks. | Can Ergen, Nir Yosef | Nature Methods · 2026-09-24 · DOI 10.1038/s41592-026-03212-9 https://developer.nvidia.com/blog/introducing-nv-reason-ct-open-3d-ct-vlm-for-radiologist-chain-of-thought-reasoning/hst-news-390Medicine, imaging & neuroscienceWed, 23 Sep 2026 22:00:00 GMTRadiology AI has made remarkable strides in detecting abnormalities across chest X-rays, pathology slides, and 2D scans. Yet one of the most clinically rich and... | NVIDIA · 2026-09-24 https://www.anthropic.com/news/claude-discovers-novel-enzyme-systemhst-news-122Proteins & molecular designWed, 23 Sep 2026 22:00:00 GMTIn early results from our new life sciences research lab, Claude agents found an enzyme system whose function is still unknown. | Anthropic · 2026-09-24 https://www.nature.com/articles/s41592-026-03225-4hst-news-1378Genomics & single-cellMon, 21 Sep 2026 22:00:00 GMTNanoTS facilitates nanopore long-read transcriptome sequencing based SNP detection and genotype calling using deep learning. | Zelin Liu, Feng Wang, Robert Wang et al. (13 authors) | Nature Methods · 2026-09-22 · DOI 10.1038/s41592-026-03225-4 https://www.microsoft.com/en-us/research/blog/improving-synthesis-prediction-of-small-molecules-at-scale-with-retrochimera/hst-news-185Drug discoverySun, 20 Sep 2026 22:00:00 GMTCustom-made molecules are advancing medicine, materials, and agriculture, but producing them is slow and expensive. A new Nature paper highlights RetroChimera, a predictive model that helps accelerate chemical synthesis, helping researchers explore a wide range of molecules. | Microsoft Research · 2026-09-21 https://www.anthropic.com/research/claude-uplifts-biomolecular-modelinghst-news-134Proteins & molecular designThu, 17 Sep 2026 22:00:00 GMTAnthropic · 2026-09-18 https://www.anthropic.com/news/life-sciences-verification-programhst-news-124Biosecurity & industryWed, 16 Sep 2026 22:00:00 GMTAnthropic · 2026-09-17 https://developer.nvidia.com/blog/high-throughput-structure-prediction-with-bionemo-inference-runtime/hst-news-411Proteins & molecular designWed, 09 Sep 2026 22:00:00 GMTBiomolecular structure prediction is now often run at proteome scale, where the goal is to move an entire worklist through the pipeline efficiently. NVIDIA... | NVIDIA · 2026-09-10 https://blog.google/innovation-and-ai/models-and-research/google-deepmind/alphagenome-atlas/hst-news-1282Genomics & single-cellMon, 07 Sep 2026 22:00:00 GMTWe’re introducing AlphaGenome Atlas, a database predicting the effects of every possible single nucleotide variant in the human genome. | Google · 2026-09-08 https://huggingface.co/nvidia/NV-Reason-CThst-news-171Medicine, imaging & neuroscienceMon, 07 Sep 2026 22:00:00 GMTNVIDIA · HF · 2026-09-08 https://research.google/blog/transfer-learning-for-genomic-prediction-in-underrepresented-populations/hst-news-118Genomics & single-cellThu, 03 Sep 2026 22:00:00 GMTGoogle Research · 2026-09-04 https://research.google/blog/a-connectomics-milestone-mapping-the-complete-male-fruit-fly-brain/hst-news-119Medicine, imaging & neuroscienceThu, 03 Sep 2026 22:00:00 GMTGoogle Research · 2026-09-04