BioAI Frontier (@BioAIFrontier) profile picture

Making sense of the frontier where AI meets biology.

Joined November 2026
Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)Photo by BioAI Frontier (@BioAIFrontier)
🧬 MIT × Sanofi: 15 antibody–antigen structure predictions pooled across five models beat 100 predictions from any single model at selecting an accurate binding pose. Their selector, AbPoolRank, reached 58.0% Success@1 versus 51.5% for the strongest single-model deep-sampling setup—without using the predictors’ native confidence scores. 🔬 Across 103 post-2021, cluster-disjoint test complexes, the team pooled three structures each from Boltz-2, Promera, Protenix-v2, ESMFold2-Fast and OpenDDE. Success meant selecting a structure with DockQ ≥ 0.49: medium-or-better docking quality, not experimental binding. 🧩 AbPoolRank ranks candidates using structural agreement, epitope overlap and source-balanced support. It knows which structures share a source, but not the predictor’s name. The main gain came from complementary candidates: different models explored useful binding poses that deeper sampling of one model missed. Effective ranking then converted that broader coverage into better selections. 🎯 The practical takeaway: allocate inference budget across complementary predictors, alongside reliable selection—not just more runs of one model. ⚠️ Preprint; computational benchmark only. Results are specific to antibody–antigen complexes. Adding another predictor did not automatically improve accuracy. The listed code repository currently returns 404. 📄 Paper: biorxiv.org/content/10.64898… #AntibodyDesign #ProteinStructure #AIforScience
45
🧬 Nature Biotechnology: RNA design agent RhoAgent used experimental feedback to iteratively improve hovlinc ribozyme cleavage efficiency, reaching up to 6× that of a human-expert-truncated construct. The ribozyme’s structure had not been experimentally characterized, making this a test of functional optimization with limited structural information. Developed by CUHK’s Yu Li group and collaborators including James J. Collins and Chun Kit Kwok, RhoAgent combines RNA generation tools with an agent-driven design workflow. 🧩 How does it work? RDiff, its diffusion-based generative tool, produces candidates conditioned on base interactions and functional motifs. RhoAgent organizes candidate design and refinement, using experimental feedback to inform subsequent iterations. The goal is to preserve or improve RNA function while exploring new sequences and structural arrangements. 🔬 The study reports several experimental demonstrations: → Diverse, active ribozymes with only 40–70% sequence similarity to their wild-type counterparts. → Hovlinc cleavage efficiency reaching up to sixfold that of the expert-truncated comparator. → Fluorogenic aptamer scaffolds and aptamer-based FRET scaffolds that spatially organize two functional cores while preserving both components’ binding activities. The contribution is the combination: constrained RNA generation, experimental evaluation and feedback-driven refinement, extending to scaffolds containing multiple functional components. ⚠️ The sixfold result is specific to hovlinc and its expert-truncated comparator—not a general improvement over wild-type RNA. This summary draws on the official abstract and public repository documentation; the full paper’s experimental details and component-level comparisons have not been reviewed. RhoAgent and RDiff code are publicly available. 📄 Paper: nature.com/articles/s41587-0… (nature.com/articles/s41587-0…) 💻 RhoAgent: github.com/ml4bio/RhoAgent (github.com/ml4bio/RhoAgent?u…) 💻 RDiff: github.com/ml4bio/RDiff (github.com/ml4bio/RDiff?utm_…) #RNADesign #SyntheticBiology #AIforScience #AIAgents #Nature Biotechnology
2
4
28
1,037
🧪 Science: a de novo bi-zinc enzyme designed without a natural template or pre-specified catalytic motif achieves a reported rate enhancement of up to 3 × 10¹³ relative to the uncatalyzed reaction—in the range of characterized natural phosphatases. Stanford, EPFL and collaborators developed dEVA, a multi-objective protein-design method, and experimentally characterized an enzyme that hydrolyzes both phosphomonoesters and phosphodiesters. 🧩 The innovation is how the catalytic site is designed. Many enzyme-design approaches begin with a predefined catalytic motif and build a protein around it. dEVA instead searches for sequences and metal environments that work together, balancing sequence–structure compatibility with predictions of a catalytically suitable zinc environment. Its evolutionary algorithm iteratively proposes changes and selects candidates representing favorable trade-offs between these objectives. 🔬 Binding a metal is not enough. The surrounding residues must position the metal and create an environment that supports chemistry. Here, the characterized design delivered hydrolytic activity, with X-ray crystallography and absorption spectroscopy supporting its de novo active site. The 3 × 10¹³ figure is defined as (kcat/KM)/kw, relative to the uncatalyzed reaction. It is not a speed advantage over natural enzymes. The contribution is a design route that reduces reliance on pre-specified active-site motifs while producing experimentally demonstrated catalysis. ⚠️ These results concern one characterized bi-zinc enzyme and the tested substrates, not a general demonstration that de novo enzymes match natural catalysts. This post draws on the Science abstract, earlier preprint and public code; the final Science full text was not accessible. Earlier versions were already public—the news is formal publication. 📄 Science: doi.org/10.1126/science.aej3… 📑 Earlier preprint: biorxiv.org/content/10.64898… (biorxiv.org/content/10.64898…) 💻 Code: github.com/gelnesr/dEVA (github.com/gelnesr/dEVA?utm_…) #ProteinDesign #EnzymeDesign #Biocatalysis #AIforScience #Science
3
13
561
🧬 In Arc Institute’s 2026 Virtual Cell Challenge, the winning predictions covered 44% of the distance between the cell context mean and an independent replicate. The organizers note that this number does not represent accuracy and that individual metrics vary widely around this average. The official wrap-up, dated November 19, 2026, reports that CellularAI Algorithm from Illumina scored 0.440 on three held-out lines. The official results table lists 0.4399, 0.3348 and 0.3263 for first through third. On this scale, 0 is the cell context mean — the average response observed across knockdowns — and 1 is an independent replicate. The overall score is the unweighted average of six metrics and three cell lines. The winners are BioInsight AI Lab from Illumina (first, $100,000), FlowState (second, $50,000) and X-team from Xaira Therapeutics (third, $25,000). Each prize is paid half in cash and half in NVIDIA Brev credits. The wrap-up lists FlowState as Independent Researchers, trained exclusively on public data, and names X-team’s model X-Cell Transfer. Zero-shot here means the Challenge provided no perturbation data from the six evaluation lines. It does not mean the models had never seen the perturbed genes. All but one of the 300 test targets already had public perturbation data in some other cell line. The final ranking used AC16 (cardiomyocyte), MOLM13 (acute myeloid leukemia) and U-87MG (glioblastoma). Validation used Cal-33, HeLa and Jurkat; the two scores are not directly comparable. All three winning teams built on reference transfer of measured responses from other cell types. They differed mainly in how they weighed the evidence, adapted it to a new line using only its control cells, and converted the predicted effects into cells: BioInsight AI Lab used a random-effects meta-analysis with weights for context similarity, noise and knockdown efficiency; FlowState learned shared response programs from a broad public corpus; X-team used a large pretrained foundation model to decide which measured effects should carry into a new context and how strongly. The organizers write that the pattern suggests that, for now, reference data and how carefully it is transferred to a new context matter at least as much as model architecture. On the validation lines, they found that the top ten entries recovered roughly what a cross-line average would predict, while the line-specific part of the response remains largely missing. ⚠️ Open weights and code were not required this year; the method descriptions are based on information provided by each team. These are computational scores on the six Challenge cell lines, not a test in primary human tissue or the clinic. 📄 Official wrap-up: arcinstitute.org/news/virtua… 🔗 Challenge site: virtualcellchallenge.org/ 📐 Scoring: virtualcellchallenge.org/eva… #VirtualCellChallenge #GenePerturbation #VirtualCell #AIforScience
2
1
6
462
🧬 AlphaGenome and Borzoi improved after adaptation to personal genomes—but still trailed a conventional elastic-net model in this study of individual-level blood gene expression. In 5,120 Hispanic/Latino participants, the authors reported mean per-gene Spearman correlations of 0.33 for adapted Borzoi, 0.29 for adapted AlphaGenome and 0.46 for elastic net. Importantly, elastic net produced no predictions for 15 of the 150 evaluated genes, so these averages do not cover identical gene sets. 🔬 The question was specific: can a model predict how expression of the same gene differs between people? Researchers from Harvard T.H. Chan School of Public Health, Beth Israel Deaconess Medical Center and collaborators paired personal genomes with whole-blood RNA-seq from HCHS/SOL. They selected 150 genes spanning low, medium and high cis-heritability—how much expression variation is attributable to nearby genetic variants—and reserved 768 participants for testing. 🧩 Adaptation helped substantially. → Borzoi: 0.11 pretrained → 0.33 after adaptation. → AlphaGenome: 0.15 → 0.29. Borzoi improved over its pretrained version for 129 genes, but exceeded elastic net for only 21. However, this was not a comparison under identical input conditions. Elastic net used variants within ±1 Mb of each gene’s transcription start site; adapted sequence models used 49,152-bp inputs. Borzoi used LoRA adaptation. AlphaGenome’s final configuration froze its backbone and trained only a regression head, after LoRA provided little additional benefit. 🌍 Ancestry analyses found broadly similar pretrained performance across European-, African- and Amerindian-enriched groups. Adapted models and elastic net generally performed slightly better when training and test ancestry groups matched. The takeaway: learning regulatory patterns from reference sequences does not automatically translate into accurate prediction of expression differences between individuals. ⚠️ Preprint; limited to bulk whole blood and 150 selected genes. This evaluates AlphaGenome, not AlphaGenome 2, and does not establish performance across all genomic tasks. The authors also considered only single-nucleotide variants when reconstructing personal sequences. 📄 Paper: doi.org/10.64898/2026.11.17.… #AlphaGenome #Genomics #GeneExpression #AIforScience
2
7
284
⚡ Rosetta physical scoring on GPUs—with up to 362× higher throughput for batched scoring and gradient calculations. tmol makes these calculations available as PyTorch components that AI protein-design workflows can call. tmol is a software library for evaluating and refining molecular structures. It implements the RosettaGenFF energy function alongside side-chain packing, energy minimization and FastRelax, helping developers incorporate physical scoring into generation and optimization. 🧩 Why does that matter? A generated structure may contain unfavorable atomic contacts or poorly satisfied interactions. tmol lets developers evaluate these features across batches of candidates. Its coordinate gradients can also indicate how moving atoms changes the energy. 🔬 The speedup has specific conditions. On an H200 NVL, batched scoring plus gradients ran 233–362× faster than single-threaded PyRosetta; torsion-space relax ran 38–102× faster. These are component-level throughput gains, not whole-workflow speedups. Across 1,000 high-resolution protein chains, total scores correlated with PyRosetta at Pearson r = 0.997. 🧪 The authors also tested physical scoring during generation. Across 24 protein–protein targets, tmol-guided RFD4-Proteina search increased the share of designs with at least five hydrogen-bonded interface residue pairs from 16% to 66%. Selecting the best 32 of 128 unguided samples already reached 53%. OpenFold3 refolding success improved from 14.5% to 22.9%, below that selection control’s 25.5%. The practical advance is making Rosetta-style energetics easier to use during AI design, with demonstrated control over specific interface features. ⚠️ Preprint; no wet-lab validation. The generation demonstration used scoring without gradients and produced larger interfaces with slightly more clashes. Benchmarks used tmol 0.2.0; v1.0.0 is available. Code is open under BSD-3-Clause. 📄 Paper: arxiv.org/abs/2611.21857 💻 Code: github.com/RosettaCommons/tm… (github.com/RosettaCommons/tm…) 📚 Documentation: rosettacommons.github.io/tmo… #ProteinDesign #Rosetta #GenerativeAI #AIforScience
5
14
857
🧬 A 36-amino-acid AI-designed protein disrupted a motor–regulator interaction and reorganized mitochondria inside cultured cells. Sidetrak, designed by University of Bristol researchers using BoltzGen, targets a regulatory surface on kinesin-1—the molecular motor that transports cargo along microtubules. The advance is a functional design application: a compact binder can interfere with a specific regulatory interaction and produce a measurable cellular phenotype. 🎯 Targeting the regulatory stalk. Sidetrak binds the TPR docking site, or TDS, on kinesin-1’s coiled-coil stalk. This surface participates in motor regulation and overlaps the binding site for MAP7, which promotes kinesin-1 recruitment to microtubules. Starting from an AlphaFold3 model, the team generated 10,000 designs. Six passed the design filters; cellular screening identified Sidetrak as the strongest candidate for further characterization. 🔬 Binding and competition were tested experimentally. → Fluorescence polarization measured an apparent KD of 110 ± 13 nM against a kinesin-1 construct containing the target site. → Disrupting either the target site or Sidetrak’s predicted interface disrupted binding. → In cell-lysate competition experiments, Sidetrak reduced MAP7-associated kinesin-1 recovery by up to approximately 85%, with half-maximal displacement at 86 nM for the heavy-chain readout. 🧫 The effect extended into cells. Cells expressing GFP-Sidetrak showed reduced MAP7-dependent motor recruitment to microtubules and a pronounced redistribution of mitochondria toward the nucleus. The binding-defective control did not reproduce the mitochondrial phenotype. The main mitochondrial analysis covered 60 cells across three independent experiments. Similar redistribution occurred in two additional cell lines, while gross microtubule architecture remained unaffected. Quantitative proteomics also supported selective engagement of endogenous kinesin-1 complexes, although off-target associations could not be excluded. 💡 Sidetrak demonstrates how designed proteins can probe regulatory interfaces and connect molecular binding to cellular function. Its interface-dependent binding also provides experimental support for the local structural model used for design. ⚠️ November 18 preprint, not peer-reviewed. Sidetrak modulates a regulatory mechanism; it is not established as a general kinesin-1 inhibitor. Cellular experiments used intracellularly expressed GFP fusions—not externally administered protein—and no therapeutic benefit was demonstrated. 🔗 Paper: biorxiv.org/content/10.64898… #ProteinDesign #Miniproteins #CellBiology #AIforScience
5
21
1,127
🧬 FoldMapper brings evolutionary information into protein inverse folding—without needing the sequence it is supposed to design. On TED domains, it achieved 60.0% best-of-100 sequence recovery versus 46.6% for ProteinMPNN under the same evaluation setting. The November 17 preprint by Haelyn Kim and Juyong Lee addresses a circular dependency: evolutionary profiles usually come from sequence alignments, but inverse folding starts with a backbone and asks which sequence should fit it. 🔍 FoldMapper starts with shape. → Retrieve structurally similar proteins using Foldseek, without a query-sequence homology search. → Aggregate their amino-acid identities into a position-specific frequency profile: which residues occur at corresponding structural positions? → Combine that profile with backbone geometry through a multiscale graph encoder and a structure-conditioned message-passing network. The model therefore uses known proteins’ sequences, but retrieves them through structural similarity rather than requiring a starting sequence for the design. 📊 Reported best-of-100 sequence recovery: → TED domains: 60.0% vs ProteinMPNN’s 46.6%. → PDB: 64.9% vs 50.3%. → Non-peptide binders: 51.3% vs 46.2%. → Peptide binders: 53.1% vs 51.0%. FoldMapper ranked first on TED and peptide binders, and second on PDB and non-peptide binders. It also sat on the recovery–diversity Pareto frontier, offering competitive trade-offs between recovering native residues and generating varied sequences. 🧩 Beyond native-sequence recovery, its predicted structure self-consistency surpassed all tested baselines on RFdiffusion-generated monomer and binder backbones, while matching the strongest baselines on natural proteins. 💡 The contribution is a sequence-independent route to retrieve evolutionary context for backbone-conditioned design. It gives inverse folding an additional source of information beyond geometry alone. ⚠️ Preprint; computational benchmarks, without wet-lab validation. Best-of-100 recovery is not single-sample accuracy or an experimental success rate. Predicted structural consistency does not establish binding or biological function. Code, checkpoints and benchmark resources are publicly available. The code is Apache-2.0; the preprint is CC BY-NC 4.0. 🔗 Paper: biorxiv.org/content/10.64898… 💻 Code: github.com/haelyn-kim/foldma… 📦 Model checkpoints: doi.org/10.5281/zenodo.24186… #ProteinDesign #InverseFolding #AIforScience #BioAI
1
11
622
🧬 Only 29 of 107 literature-reported AI nanobody binders showed binding in Spiritein’s retest—a 27.1% hit rate. Separate design campaigns yielded 9 binders from 221 IL-7Rα candidates and none from 105 PD-L1 candidates. Spiritein’s assessment asks a practical question: how well do published binding results and computationally selected designs hold up in a standardized experimental pipeline? 🔬 Published binders: 29 positives out of 107. Spiritein retested literature-reported sequences through standardized expression, purification and characterization. Seventy-eight showed no detectable binding under the retest conditions. An important qualification: among 27 confirmed binders with paired KD measurements, the report found no significant overall difference between published and retested affinities. The main discrepancy was whether binding was detected—not uniformly weaker affinity among confirmed binders. 🧪 New designs: approximately 1.16 million sequences generated. For IL-7Rα, the team compared 13 workflows combining structure-generation and sequence-design methods. Approximately 960,000 sequences underwent computational screening, with final ranking by AlphaFold3 ipTM. A third-party CRO tested 221 selected candidates: → Nine binders overall: a 4.1% hit rate. → Eight came from BoltzGen paired with ProteinMPNN or SolubleMPNN. Each route yielded 4/20 hits; the best KD across the two routes was 8.82 nM. → One came from PPIFlow paired with SolubleMPNN. For PD-L1, four workflows—BoltzGen, Germinal, OpenDDE-harness and HalluDesign—yielded no detectable binding hits among 105 candidates, despite nearly all being successfully expressed and purified. 🧩 Binding was only the first checkpoint. Confirmed binders were also assessed for thermal stability, heat-induced aggregation and binding to baculovirus particles, an indicator of nonspecific interactions. Results varied widely, and some candidates showed elevated nonspecific-binding signals. 🧭 The report draws three conclusions: → Reproducibility: some previously reported high-affinity binding results could not be reproduced under this round’s testing conditions. → Preclinical readiness: binding alone does not establish a development candidate. Binding kinetics, stability, aggregation and specificity remain important hurdles. → Performance on established targets: current AI workflows still struggle to reliably deliver experimentally validated binders on targets with mature antibody-development pathways, as illustrated by the PD-L1 results. ⚠️ These conclusions come from a company-reported assessment under specific design, selection and assay conditions. They should not be treated as universal model failure rates or proof that all original experiments were invalid. Spiritein says it plans to release the code, workflows and raw experimental data through AbGround, supporting closer examination and independent reproduction of the results. 🔗 AbGround / Spiritein: spiritein.com/ #AntibodyDesign #ProteinDesign #AIforScience #BioAI
2
15
936
🧬 Anthropic × BigHat brought Claude into antibody optimization to tackle a blind spot in automated design: a model can improve an experimental score without improving the property researchers actually need. Across two campaigns, researchers and Claude examined raw instrument curves, uncovered misleading thermal-stability metrics and tested revised interpretations in new experiments. Their preprint, “Reasoning in the loop,” extends the optimization loop beyond choosing better molecules to questioning—and experimentally checking—how “better” is measured. 🔬 Seven rounds, approximately 1,010 design variants. On BigHat’s Milliner platform, the team optimized an anti-albumin nanobody and an anti-mesothelin scFv, a linked antibody fragment. Claude helped investigate antibody lineage, structural evidence and experimental curves, and propose hypotheses for testing. Researchers set the strategy, reviewed designs and performed the experiments. 📉 Two campaigns exposed two measurement problems: → Nanobody: aggregation began around 40°C, well below its melting temperature. Raising the melting point did not necessarily reduce aggregation. → scFv: the reported apparent “melting temperature” reflected protein precipitation rather than a clean unfolding transition. The team fitted physical models to the curves, adopted more suitable metrics, and varied protein concentration and heating rate to test preregistered predictions. 📊 Both final leads improved over their in-campaign parents: → Nanobody: KD improved from 10.3 to 1.15 nM—about 9× tighter binding. Confirmed melting temperature rose from 62.2°C to 70.6°C, with no aggregation onset detected up to the 85°C measurement limit. → scFv: KD improved from 3.7 to 0.63 nM—about 6× tighter binding. Solution-collapse temperature rose from 56.3°C to 74.2°C, approximately +18°C. 💡 The contribution is bringing scientific reasoning about measurement into a live optimization campaign. The agent helped researchers investigate why a readout behaved unexpectedly, formulate a physical explanation and test it—alongside designing improved molecules. ⚠️ Preprint; two human-directed lead-optimization campaigns. Lead measurements had limited replication, some affinities approached the assay’s resolution limit, and not all preregistered predictions were fully confirmed. The results do not isolate Claude’s contribution or establish autonomous drug development. 🔗 Paper: arxiv.org/abs/2611.19812 (arxiv.org/abs/2611.19812?utm…) #AntibodyEngineering #AIforScience #LabAutomation #BioAI #Anthropic #Claude
1
3
17
823
💰 Alphabet’s Isomorphic Labs has raised $5 billion in a single round—bringing its announced funding to $7.7 billion in roughly 20 months, while its first drug candidates are still heading toward human trials. The next test: whether its AI drug-design engine can turn computational gains into medicines that work in patients. Announced on November 17, the Series C was co-led by Greenoaks and Sequoia Capital, with a16z and other new investors joining. Alphabet invested again, alongside all existing backers: Thrive Capital, MGX, GV, CapitalG, Temasek and the UK Sovereign AI Fund. The final valuation was not disclosed. 📈 Three rounds, with financing accelerating: → March 2025: $600 million in its first external funding round → May 2026: $2.1 billion Series B → November 2026: $5 billion Series C That is $7.1 billion raised in approximately six months. 🧬 What is the money scaling? Founded in 2021 and led by Demis Hassabis, Isomorphic is building IsoDDE—the Isomorphic Labs Drug Design Engine. It combines predictive models to help researchers understand molecular interactions, select candidates for synthesis and testing, and refine their properties. Its February technical disclosure described three capabilities that extend beyond predicting a protein’s structure: → Structure prediction: more than twice AlphaFold 3’s accuracy in the hardest generalisation category of Runs N’ Poses, involving unfamiliar protein pockets and ligands. → Binding affinity: predicting how strongly a molecule binds, helping researchers rank and optimise candidates. → Pocket identification: finding potential binding sites from protein sequences, including hidden pockets that are difficult to identify in an unbound structure. 🤖 The latest announcement puts proprietary AI agents alongside these models. Isomorphic says the agents reason over complex biological systems, supporting its drug-development programmes. It reports positive preclinical results in internal oncology and immunology projects and partnered programmes. The new capital will also fund translational research and clinical-development infrastructure, supporting the work needed to move candidates from the laboratory into human testing. 💡 My read: investors are backing the possibility of a reusable drug-development platform that can support many programmes. Collaborations with Lilly, Novartis and Johnson & Johnson provide real projects on which to test its practical value. ⚠️ Clinical validation remains ahead. Accurate binding predictions do not establish human safety or efficacy, and positive preclinical findings do not guarantee clinical success. IsoDDE’s published benchmarks are company-reported; this announcement does not report clinical entry. The milestones to watch now: which candidate reaches human testing first, what disease it targets, and whether the resulting data support the promise of the platform. 🔗 Official funding announcement: prnewswire.com/news-releases… (prnewswire.com/news-releases…) 🔗 IsoDDE technical overview: isomorphiclabs.com/articles/… (isomorphiclabs.com/articles/…) #AIDrugDiscovery #AIforScience #Biotech #Alphabet #Isomorphic Labs
3
7
480
🧬 Claude Fable 5.2 outperformed the tested open protein language models—including ProGen3-3B—at predicting the next amino acid in little-studied proteins. An external analysis points to substantial protein-sequence learning in Claude. In a November 16 analysis, Xinming Tu and Leo Tianlai Chen investigated a question behind LLM performance on protein tasks: do these models learn sequences themselves, or mainly recall descriptions from papers and webpages? 🧪 The test: unfamiliar proteins, with shuffled controls. The researchers selected 668 bacterial and archaeal proteins from 167 Pfam families of unknown function. These families had no experimentally resolved structures and at most three associated papers each. Each protein was cut at five positions, creating 3,340 prediction tasks. Models received a sequence prefix and generated ten amino acids; the primary metric scored only the first. They then repeated the test with shuffled sequences, preserving amino-acid composition while destroying sequence order. 📊 Real-sequence accuracy minus shuffled-sequence accuracy: → Claude Fable 5.2: +45.1 percentage points → ProGen3-3B: +32.5 → ProtGPT3-10B: +27.7 → Claude Opus 5.5: +27.4 Fable’s accuracy was 54.1% on real sequences versus 9.0% on shuffled controls. The tested GPT, Gemini and seven open-weight models showed little comparable separation. 🔍 Three findings strengthen the sequence-learning interpretation: → The effect persisted across 128 families with no associated papers, weakening the explanation that models simply recalled published descriptions. → Performance increased with protein-family size—a pattern also observed in dedicated protein language models. → High accuracy did not depend on reported reasoning tokens, suggesting that extended deliberation was not driving the result. The signal extended beyond one residue: Fable correctly continued all ten amino acids at 88 cut points; Opus did so at 12. Both scored zero on shuffled controls. 💡 Some general-purpose LLMs may already contain substantial protein-sequence knowledge. Whether that knowledge can support useful mutation-effect prediction, folding or functional design remains an open experimental question. ⚠️ Public analysis with open code; no peer review or wet-lab validation. This benchmark tests sequence continuation, not overall protein-design capability. Prefix-only inputs disadvantage bidirectional ESM models, and the inferred training exposure has not been confirmed by Anthropic. 🔗 Original analysis (xinmingtu.cn/blog/2026/claud…) · Code and results (github.com/XinmingTu/LLM-bio…) #AIforScience #ProteinLanguageModels #BioAI #Claude #ClaudeFable5.2
1
8
26
1,200
🧬 Nature Medicine has published OpenScientist, an open-source AI co-scientist evaluated across five biomedical case studies—including hidden-data testing, repeated runs and blinded comparison with independent human analysis. The study shows both the research workflows it can execute and the errors that still require expert correction. One concrete result: a six-protein survival score achieved a c-index of 0.826 in a hidden cohort of 176 participants, including 17 deaths. The model was fixed before evaluation on that cohort, testing whether it could rank survival risk in people whose data had not been used to build the score. Here’s what makes the paper worth reading: → It executes analyses, beyond proposing ideas. Given a scientific question and data, OpenScientist can search literature, write and run code, inspect results, iterate and produce a report with analysis artifacts available for review. → The evaluation goes beyond a single successful run. Across different cases, the authors examined performance on unseen data, repeated a tau-pathology analysis across 60 runs under different configurations, and compared AI and independent human proteomic analyses under a blinded, prespecified setup. → The failures are consequential. The evaluations identified issues involving duplicate observations, cohort definitions and statistical reporting. In one case, a stricter statistical correction was calculated but omitted from the final report. Executable code alone does not guarantee a sound conclusion. → Sensitive data can stay offline. The system is open source under Apache 2.0 and supports fully air-gapped operation with local open models—a useful option for clinical research environments. The contribution is a framework for executing and inspecting biomedical analysis workflows, with concrete tests of reliability. It also makes the remaining gap clear: generating an analysis and establishing a trustworthy conclusion are separate achievements. ⚠️ The survival result comes from one case with only 17 deaths in the hidden cohort; c-index 0.826 does not mean 82.6% diagnostic accuracy. These are computational biomedical studies, not autonomous wet-lab experiments or evidence of clinical readiness. 📄 Paper: nature.com/articles/s41591-0… 💻 Code: github.com/openscientist-io/… #AIAgents #AIforScience #ComputationalBiology #OpenSource
4
12
607
🧬 David Baker and Anthony Green’s teams designed enzymes from scratch to hydroxylate different sites on the same molecule: estradiol. Separate designs targeted C6 or the unactivated C16 position, with 93% and >99% diastereomeric excess, respectively. The starting point was a predefined substrate and reaction geometry. The team then built new protein scaffolds around the catalytic machinery needed to perform the reaction. In this bioRxiv preprint, researchers at the University of Washington / HHMI and the University of Manchester report: → Reaction geometry as the design input. A cysteine-ligated heme, catalytic residues for hydrogen peroxide activation, and the target C–H bond positioned beside the reactive iron–oxo intermediate defined the active site. RFdiffusion2/3 generated scaffolds, followed by sequence design and all-atom filtering. → Different sites on the same complex substrate. The estradiol designs demonstrate control over both the position of hydroxylation and the stereochemistry of the product. The reported 93%/>99% d.e. values describe stereoselectivity—not conversion or yield. → Activity improved through further engineering. In the C16 route to estriol, substitutions increased yield from approximately 1% to 37% and total turnover number from below 1 to 37, while retaining >99% d.e. In a separate aliphatic-substrate campaign, an improved variant reached 250 turnovers, versus 12 for its parent design. → Crystal structures supported the designs. A 1.9 Å structure of an estradiol hydroxylase variant matched its AlphaFold3 model with a Cα RMSD of 0.65 Å. The advance is a concrete route toward custom biocatalysts: specify the substrate and target reaction geometry, then construct a protein around them. That could expand options for selective modification of complex molecules. ⚠️ Preprint with wet-lab and crystallographic validation. The demonstrated substrates and turnover numbers do not establish an industrially ready catalyst or the ability to target any C–H bond on demand. Further optimization was needed. 📄 Paper: biorxiv.org/content/10.64898… #ProteinDesign #EnzymeEngineering #Biocatalysis #AIforScience #SyntheticBiology
3
28
125
6,724
🧬 In Nature Communications, an mRNA vaccine encoding two computationally designed influenza hemagglutinins gave mice 100% survival in five of six lethal influenza challenges—and 60% survival against H10N8. The work introduces PanFlu-IOA, a computational antigen-design toolkit built to extract shared antigenic features across influenza A Group 1 and Group 2. 🔬 The team designed two universal HA sequences: → G1-HA for Group 1 → G2-HA for Group 2 The two mRNA-LNPs were mixed into the vaccine candidate panFluA. Two weeks after the booster, female BALB/c mice were challenged with 10 LD50 of six divergent viruses. Reported survival: → 100%: pandemic H1N1, seasonal H1N1, PR8, H9N2, and H3N2 → 60%: H10N8 → Controls showed near-complete mortality by days 7–9 T-cell depletion and passive-serum-transfer experiments supported contributions from both cellular and humoral immunity. PanFluA serum alone provided about 80% protection against PR8. Why it matters: this connects cross-group computational HA design to multi-subtype lethal-challenge protection in mice—an end-to-end design-to-challenge validation rather than a purely computational antigen study. ⚠️ This is a mouse study, not a human vaccine result. Cross-protection was not uniformly matched by classical neutralizing or hemagglutination-inhibition activity outside H1 strains, so protection cannot be attributed to those antibody readouts alone. Recent isolates from the past five years were not tested, and pre-existing human immunity was not modeled. 📄 Paper: nature.com/articles/s41467-0… 💻 Code: github.com/xielisos567/PanFl… #Influenza #mRNA #VaccineDesign #ComputationalBiology #AIforScience
6
15
681
🧬 MDClaw found 11 AI-run MD workflows that looked successful—but had skipped minimization or equilibration, or simulated the wrong structure. That is the failure mode this new system is designed to address: an agent can produce a trajectory and a convincing report without actually following the intended molecular-dynamics protocol. MDClaw is not another MD foundation model. It is an auditable constraint layer for general coding agents: validating command-line tools check inputs and stage dependencies before execution, while a directed acyclic graph (DAG) records every stage in a recoverable, branchable workflow. 🔬 On MDDataBench, a 31-system benchmark scored against deposited reference simulations: → Without MDClaw: 21 passes in 372 independent runs, or 5.6% pass@1 → With MDClaw tools: 81–94% pass³ across four open-weight models → With tools plus skills: 90–100% pass³ Here, pass³ means a system passes in all three independent runs—not merely one lucky successful attempt. Why it matters: as agents begin to simulate atomic-scale scientific hypotheses, the important question is not only “did it produce a trajectory?” It is whether a scientist can inspect what was actually simulated, which inputs were used, and whether essential stages were skipped. ⚠️ This is a bioRxiv preprint. MDDataBench measures correct execution against predefined simulation tasks and reference outputs; it does not establish sampling convergence or biological correctness. The main results cover four specific open-weight models in one agent setup. Its no-reference applications are traceable simulation studies, not experimentally confirmed biological discoveries. 📄 Preprint: biorxiv.org/content/10.64898… 💻 Code: github.com/matsunagalab/mdcl… 🧪 Benchmark: github.com/matsunagalab/MDDa… #MolecularDynamics #AIAgents #ScientificAI #ComputationalBiology #Reproducibility
7
36
1,400
🧬 No experimental structure. About 3 million compounds screened. One AlphaFold-guided campaign produced QL07, a CLIP1–LTK inhibitor with an enzymatic IC50 of 524.27 nM. CLIP1–LTK is an oncogenic fusion kinase without an experimentally resolved structure. In a new Acta Pharmacologica Sinica paper, researchers first benchmarked how AlphaFold2- and AlphaFold3-derived structures perform in virtual screening, then applied a hierarchical workflow to this target. 🔬 After screening approximately 3 million compounds and conducting similarity-based analog searching, they identified QL07. The study reports that QL07: → inhibited downstream CLIP1–LTK signalling → induced apoptosis in Ba/F3 cells transformed with CLIP1–LTK → suppressed growth of CLIP1–LTK-positive xenografts Why it matters: this is not only an in silico docking exercise. It is a peer-reviewed wet-lab case study showing that AlphaFold-derived structures can help start a structure-based discovery campaign when a disease target has no experimentally determined structure. ⚠️ QL07 is an early research compound, not a clinical candidate. The paper does not establish superiority over existing LTK or ALK inhibitors, and its reported xenograft activity does not demonstrate human efficacy. No public code or model-weight release was stated in the abstract. 📄 Paper: nature.com/articles/s41401-0… #AlphaFold #DrugDiscovery #VirtualScreening #CancerResearch #AIforScience
2
5
13
991
🧬 Given only the sequence of the EBV protein BHRF1—no structural template—LiteFold’s LiteMol-1 selected 19 linear peptides for testing. Nine bound by BLI. That is a 47.4% hit rate in the selected set. Two leads showed apparent KDs of 37.6 nM and 49.9 nM, lower than the natural ligand Bim at 75.6 nM in the same assay format. 🔬 LiteMol-1 is a masked diffusion language model for target-conditioned peptide generation. The workflow generated and cofolded roughly 500–600 candidates, then used Protenix v2, Boltz-2, interface scoring, and manual selection before sending 19 peptides for capture-format BLI testing at Adaptyv Bio. Results: → 9 BLI-positive peptides → 9 BLI-negative peptides → 1 expression failure → selected peptides were 18–27 residues long Why it matters: this is a useful early wet-lab readout for a sequence-conditioned peptide-generation model—not merely a structural prediction or in silico ranking result. ⚠️ This is a company technical report, not peer-reviewed research. The 47.4% is the hit rate after computational filtering, not the raw generation success rate. The reported values are apparent KDs from tagged capture-format BLI, not solution KDs; all nine hits carried N-terminal tags. There are no selectivity, binding-site, cellular-activity, or stability data, and no head-to-head comparison with other generators. LiteMol-1 weights and code have not been announced as public. 🔗 Technical report: litefold.ai/research/bhrf1 𝕏 Official post: teapaw.org/try_litefold/status/21… #PeptideDesign #ProteinDesign #GenerativeAI #BLI #AIforScience
We experimentally tested 19 AI-designed peptides. 9 showed binding. 2 achieved lower apparent KD than a natural ligand. Today, we're sharing the first wet-lab validation results of LiteMol-1, our molecular diffusion language model. We targeted BHRF1, an anti-apoptotic protein encoded by Epstein–Barr virus, which is implicated in certain lymphomas. LiteMol-1 generated linear peptides conditioned only on the target sequence. We then prioritized candidates using Boltz-2, Protenix along with interface scoring. We collaborated with @adaptyv independently tested the selected peptides using biolayer interferometry (BLI). The results: > 9/19 candidates showed binding (47.4%). > Peptide_010: 37.6 nM apparent KD. > Peptide_012: 49.9 nM apparent KD. > Natural Bim control: 75.6 nM apparent KD. These are early, single-target biochemical results, but an important step toward connecting generative molecular design with experimental validation. Our broader goal is to close the loop between AI-driven generation, computational evaluation, and real-world experimental feedback. Technical report link in the comments.
14
918
🧬 On 25 held-out computational biology tasks, the top AI agent scores just 40.3/100. FrontierBioML, released by Valen Labs, evaluates agents across antibody engineering, protein engineering, structure, binder design, and functional genomics. Agents write prediction scripts in a GPU-free sandbox with roughly six hours of runtime; submissions are then scored on held-out data. 🏆 Current leaderboard: → Claude Mythos 5.2 (Claude Code): 40.3 → Gemini 4 Carbon: 39.0 → Claude Opus 5.5: 36.5 → Baseline: 13.1 The important context: scores are normalized per task. The benchmark authors say 0 usually corresponds to a simple baseline, 100 to a perfect answer, and a score of 30 is often already strong. Scores therefore should not be read as a conventional “percentage correct.” No system solves all 25 tasks. The result is less “Mythos wins again” than a useful reality check: even leading agent setups remain far from reliable general-purpose computational biologists. ⚠️ This is a company-run benchmark and compares full model-plus-scaffolding systems, not base models in isolation. Results are not peer reviewed, and task scores are not directly comparable across tasks. Two GPT systems were also affected by biosecurity filtering: zero-scored blocked runs lower their overall averages, while their scored-only runs were reportedly closer to the leading group. 🔗 Benchmark: frontierbioml.com/ 📄 Details: valenlabs.com/blog/frontierb… #AIforScience #ComputationalBiology #ProteinEngineering #AIAgents #Bioinformatics
2
4
14
740