{"data":{"event":{"id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","slug":"lucacell-a-sequence-centric-foundation-model-for-cross-species-single-ce-ecc7246592","title":"\n\nLucaCell: a sequence-centric foundation model for cross-species single-cell analysis \n\n","short_summary":"Single-cell foundation models have transformed transcriptomic analysis, yet most rely on fixed gene identifiers that limit transfer across species and data types. Here we present LucaCell, a sequence-centric foundation model that represents genes through pre-trained mRNA sequence embeddings rather than static gene annotations. Gene expression is discretized into bins and modeled with a Transformer encoder, enabling sequence-informed cell representation without a fixed gene-ID vocabulary. Pre-training on 85 million human and mouse single cells, LucaCell is evaluated on human, mouse and lemur ge","full_description":"Single-cell foundation models have transformed transcriptomic analysis, yet most rely on fixed gene identifiers that limit transfer across species and data types. Here we present LucaCell, a sequence-centric foundation model that represents genes through pre-trained mRNA sequence embeddings rather than static gene annotations. Gene expression is discretized into bins and modeled with a Transformer encoder, enabling sequence-informed cell representation without a fixed gene-ID vocabulary. Pre-training on 85 million human and mouse single cells, LucaCell is evaluated on human, mouse and lemur gene expression profiles, human chromatin accessibility data, unaligned reads from more than 50 prokaryotic taxa, and five influenza A virus genomes. LucaCell enables manual-mapping-free cross-species cell type annotation and an alignment-free microbial embedding framework that simultaneously distinguishes bacterial species identity and intra-species physiological states. It also improves gene expression reconstruction by incorporating donor-specific exonic SNP information into mRNA sequence embeddings, and predicts cellular viral load across influenza A virus strains while highlighting infection-like transcriptional states in mock-infected cells. These results show that sequence-informed gene representation can improve the generalization of single-cell foundation models across species, data types, and predictive tasks.\n\nUpdate (2026-09-16): \n\nEstimand-aware and donor-aware triangulation of predefined gene-set signals across human tendon transcriptomic datasets \n\n. We analysed five predefined programmes across three human tendon transcriptomic contexts: acute exercise-associated fibroblast pseudobulks (E-MTAB-15400; 4 control and 4 exercise samples), mechanically stretched donor-derived cells (GSE150482; 3 normal and 3 tendinopathy donors), and paired lesional versus grossly normal-appearing tendon (GSE26051; 23 donor pairs). Inference was conducted at the sample, donor or donor-pair level. Non-stimulated GSE150482 single-cell RNA sequencing comprised only N1 and D1, whereas stretched single-cell RNA sequencing comprised N1-N3 and D1-D3; the conditions also used Drop-seq and 10X Chromium, respectively. A donor-replicated disease-by-stimulation interaction is therefore not estimable. Within-context Benjamini-Hochberg adjustment supported acute mechanical-response, extracellular-matrix (ECM) and protein-folding differences and paired-lesion collagen, integrin and ECM differences, but no stretched donor-background difference. Exact label permutations, a conservative global 15-test adjustment, leave-\n\nUpdate (2026-09-18): \n\nMapping Gene Expression to an Interpretable Semantic Space \n\n. Cell embeddings organize single-cell expression data, but their dimensions have no biological meaning, so clusters are interpreted afterward. We present MESIC (Mapping Expression to Semantic space with Interpretable Components), which builds the written knowledge about genes held in curated databases into the dimensions themselves. A biomedical language model converts each gene's summary into a semantic embedding. MESIC compresses these embeddings into a small number of components, each concentrated on a small set of genes and explained by their annotations. The components are computed once from the summaries, so any expression dataset can be mapped onto them, and every cluster, outlier, or cell-type assignment is then characterized by named genes. In cardiomyocytes, outliers in the component space were enriched for hypertrophic cardiomyopathy. In a lung atlas, unsupervised clusters in that space matched the broad cell types that experts had annotated. In both, the components that separated the cells matched their known biology. For about half of the cells that the atlas itself had left unannota\n\nUpdate (2026-09-18): \n\nPretrained gene representations transfer mean expression more broadly than spatial patterns in virtual spatial transcriptomics \n\n. Models that combine tissue images with pretrained gene representations aim to predict spatial expression for genes not used to fit the downstream predictor. Yet success on held-out genes can reflect two capabilities: estimating a gene's mean expression across tissue locations and recovering its spatial variation. Across four cohorts spanning three human brain regions and HER2-positive breast cancer, we evaluated held-out genes in held-out individuals and separated these components. For spatial predictors using fixed gene representations from Decima or scGPT, reductions in gene-mean error accounted for more than 91% of the reduction in mean squared error relative to matched random vectors. Independently fitted mean-only models using the same representations but no tissue images retained 90-99% of the corresponding gain in full-matrix correlation. Spatial gains were smaller on average, increased with expression variation in training tissue and differed across cohorts and representations. Across these settings, pretrained gene repr\n\nUpdate (2026-09-23): \n\nMIRCID: Inferred Hub-miRNAs Drive Cross-Task Improvements in Drug Mechanistic Modeling \n\n. Drug mechanism-of-action (MoA) modeling commonly relies on perturbational transcriptomes, but matched microRNA (miRNA) measurements are often unavailable. Inferred regulatory features offer a scalable way to reuse these data. Here, we present MIRCID, a framework comparing gene expression with inferred transcription factor (TF) activity and miRNA expression across pathway classification and similarity-based MoA retrieval. HubmiRNet infers 414 pan-cancer hub miRNAs (HubmiRs) from 977 L1000 landmark genes, achieving a Pearson correlation coefficient of 87.72%; its 1,298-output variant also outperformed SiCmiR on the full-miRNA task (71.21% versus 67.30%). In the evaluated comparisons, miRNA augmentation provided more consistent gains than TF activity. Generic embedding controls showed model-dependent utility, while complementarity analyses identified a distinct, partially linearly recoverable representation that retained gene-derived structure. Illustrative rescue cases linked improved classification to biologically plausible miRNA patterns in samples with weak transcrip\n\nUpdate (2026-09-24): \n\nISP Platform powered by Geneformer: Framework for Cross-Species, Sequential, and Multi-Gene In Silico Perturbation Screens with Application to iPS Cell State Transitions \n\n. In silico perturbation (ISP) enables virtual genetic screens, prioritizing candidate regulators before costly wet-lab experiments. However, applying foundation models requires programming expertise. Here, we present the ISP Platform, that automates ISP powered by Geneformer without code execution. The platform enables cross-species analyses by independently selecting input species and human or mouse Geneformer models. The platform also implements sequential multi-gene ISP, in which each perturbation chains from the cell state produced by the preceding step. Using induced pluripotent stem cell (iPSC) datasets, the platform successfully performed cross-species analysis of human and mouse single cell-RNA (scRNA)-seq data. Screening all 24 permutations of the Yamanaka factors identified an optimal overexpression sequence that produced the highest shift toward the pluripotent state. Genome-wide knockdown screens prioritized key pluripotency regulators but revealed an asymmetric cross-species\n\nUpdate (2026-09-24): \n\nFrom single-sequence structure prediction to protein fitness landscape through a composable, epistasis-aware mutation atlas \n\n. Mapping the multi-mutant fitness landscape is vital to protein engineering, but is challenging due to the vast combinatorial sequence space awaiting exploration. A central difficulty lies in the accurate and efficient modeling of non-additive epistatic effects among individual mutations, which partially arise from the physical inter-residue interactions prescribed by the protein structure. Existing fitness predictors usually perform well on single mutants but become less powerful for higher-order mutants, due to the lack of explicitly considering the relationship between sequence, structure and function of the target protein. Here, we present an end-to-end framework named Cerebra-Epistasis, which couples a single-sequence structure predictor that explicitly endows the structure awareness beyond the conventional sequence-fitness mapping with a downstream fitness prediction network that deliberately models the non-linear epistatic effects beyond the traditional additive terms. When evaluated across diverse assays, Cerebra-Epistasis o\n\nUpdate (2026-09-24): \n\nASOFormer: a Transformer-based model for predicting antisense oligonucleotide efficacy to support therapeutic candidate prioritization \n\n. Antisense oligonucleotides (ASOs) represent a promising therapeutic modality for RNA-based disease mechanisms, but identifying high-efficacy candidates from large sequence spaces remains a major bottleneck in drug development. Here, we present ASOFormer, a novel Transformer-based neural network that predicts the inhibition efficiency of RNase H-mediated ASOs from sequence, predicted secondary structure, and chemical modification features. ASOFormer was trained on a knockdown efficacy dataset spanning over 170,000 ASO-target pairs compiled largely from published patents. Ablation analysis confirmed that the addition of chemical modification features to the primary sequence drove the largest performance gains, with predicted secondary structure providing complementary benefit when combined with modification information. Further analysis highlighted the importance of the wings of gapmer ASOs and other features. As independent validation, ASOFormer was applied to ASOs targeting seven genes that were not in the training datas\n\nUpdate (2026-09-24): \n\nBoltz2-Notebook: An Interactive Google Colab Platform for Diffusion-Based Biomolecular Structure and Binding Affinity Prediction using the Boltz2 model. \n\n. Recent advances in deep learning-based structure prediction, including AlphaFold3 and the open-source Boltz model family, have extended biomolecular modeling to joint prediction of protein-ligand, protein-nucleic acid, and multi-chain complexes with binding-affinity estimation. Boltz-2 is among the most feature-complete of these open models, but its practical use requires a local CUDA-capable GPU, command-line execution, and manually authored YAML configuration files, limiting accessibility for researchers without dedicated computational infrastructure. We developed Boltz2-Notebook, a Colab-native interface comprising four integrated stages - automated environment setup, interactive parameter-to-YAML generation, execution management, and automated confidence and affinity visualization - together with a manifest-driven batch mode for multi-target screening. All modelling capabilities are inherited unmodified from Boltz-2; Boltz2-Notebook's contributions are limited to accessibility, input construction, a\n\nUpdate (2026-09-24): \n\nSurfGraphPro: Integrating Protein Language Models with Geometric Deep Learning On Coarse Protein Surfaces for Binding Site Prediction \n\n. Accurate and fast prediction of protein binding sites remains essential for understanding molecular interactions and facilitating protein engineering. Here, we present SurfGraphPro, a geometric deep learning approach that bridges protein language models with surface-based structural representations for a rapid binding interface prediction. Our method operates on coarse-graphed triangulated protein surfaces, utilizing solvent-excluded surface meshes downsampled into amino acid residue centered patches. Unlike existing approaches that rely on extensive physicochemical feature engineering, we leverage evolutionary information directly from protein language model embeddings, eliminating the need for computationally expensive multiple sequence alignments and hand-crafted features. This integration achieves on average $sim18-28 times$ speedup for proteins of $sim$100 to a few 1000s of amino acids over current state-of-the-art surface-based model while maintaining comparable accuracy on diverse binding interfaces, including chal\n\nUpdate (2026-09-24): \n\nDe novo design of small cyclic peptide inhibitors from non-canonical amino acids by cofolding-guided search \n\n. Cyclic peptides can engage targets beyond the reach of small molecules, but their discovery typically relies on libraries reused across targets, diversification around a known interaction motif, or computational methods limited to a handful of non-canonical building blocks. Here, we present nCycle-Forge, a simulated annealing framework that designs cyclic peptide binders by iteratively substituting building blocks and scoring each proposal by cofolding with the target. We designed, synthesized, and screened 1,472 cyclic peptides against the active-site pocket of Thrombin and the shallow p53-binding surface of MDM2 from the target sequence alone and without fixing binding mo-tifs. The library yielded chemically diverse hits against both targets in a single round without experimental optimization, with hit rates comparable to or up to 24-fold higher than in larger experimental libraries. Purified designs reached sub- to low-micromolar potency, with the best Thrombin inhibitor at Ki 0.14 M and the best MDM2 binder at EC50 2.4 M. Surface plasmon reson\n\nUpdate (2026-09-24): \n\nA Statistical Guideline for Analyzing Repeated-Measures Proteomics Data \n\n. Mass spectrometry-based proteomics increasingly supports study designs with repeated measurements, such as longitudinal sampling and spatial tissue profiling. These designs offer rich biological insights but introduce statistical challenges due to within-subject correlations and variance heterogeneity across time points or anatomical locations. Despite the growing use, best practices for analyzing repeated-measures proteomics data remain limited. We systematically compare commonly applied methods, including paired t-tests and linear mixed-effects models with correlation induced by random effects, to more flexible covariance-pattern models with unstructured covariance. Through simulation studies, we evaluate false-positive rates and statistical power under varying conditions of covariance heterogeneity. Our results demonstrate substantial limitations of conventional approaches and highlight the robustness and superior performance of unstructured covariance pattern models for repeated measures proteomics. This approach accommodates diverse study designs supporting accurate inference a\n\nUpdate (2026-09-24): \n\nAn interpretable open platform for sequence-based antibody developability prediction \n\n. Antibody developability is increasingly predictable from sequence, yet software and trained models are rarely made available. We present DELPHI, open software for training developability predictors from labelled antibody assay data, together with ready-to-run, retrainable models. DELPHI compares 25 language-model and classifier combinations under CDR H3-cluster cross-validation that reduces sequence-similarity leakage, measures how performance changes with labelled training-set size, and reports residue-level model attributions. Applied to in-house polyreactivity and size-exclusion (SEC) data, it reaches mean AUC 0.959 and 0.933. Trained on those data alone, it transfers to a 246,293-antibody public library (AUC 0.950 with our deployed model) and ranks polyreactivity at a level similar to the best reported Ginkgo competition point estimate, without training on its data. Any laboratory can screen candidates before running assays, generate residue-level engineering hypotheses, and retrain DELPHI for a new assay.","ledger_type":"benefit","primary_domain_id":"8f1af1b9-7302-4b51-aae9-fdfac04d158a","event_status":"provisional","event_date":"2026-09-14T00:00:00.000Z","discovery_date":"2026-09-14T00:00:00.000Z","first_published_date":"2026-09-14T00:00:00.000Z","last_reviewed_date":"2026-09-24T00:00:00.000Z","geographic_scope":"International","affected_population":null,"base_impact_tier":1,"base_score":"1.00","attribution_multiplier":"0.1000","evidence_multiplier":"0.5000","realization_multiplier":"0.2000","durability_multiplier":"0.5000","current_event_score":"0.005000","confidence_level":"low","score_explanation":"Ingest update: independent source raised evidence; added a follow-up source. Score remains provisional until review. Follow-up ingest: independent source raised evidence; added a follow-up source. Follow-up ingest: independent source raised evidence; added a follow-up source. Follow-up ingest: independent source raised evidence; added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source.","methodology_version_id":"8745541b-0f8e-4d5a-9b47-5291d3813a8d","original_methodology_version_id":"8745541b-0f8e-4d5a-9b47-5291d3813a8d","published_at":"2026-09-15T03:00:54.393Z","created_at":"2026-09-15T03:00:54.393Z","updated_at":"2026-09-25T00:00:55.627Z","flags":[],"domain_name":"Biology","domain_slug":"biology","methodology_version":"0.2"},"contributions":[{"id":"531298e7-e03c-4555-941a-a96c11931136","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":"535014f7-c941-4d39-aa9b-8f9c1d3d07b7","role_description":"Unspecified system mentioned or implied by a news item. Remap to a named release when identified.","attribution_multiplier":"0.1000","credit_share":"1.0000","contribution_score":"0.005000","attribution_rationale":"News ingest does not infer a named model from the publisher alone. Attribution stays unspecified until a release is identified.","attribution_confidence":"medium","first_used_date":"2026-09-14T00:00:00.000Z","model_version_if_known":null,"review_status":"approved","created_at":"2026-09-15T03:00:54.409Z","model_slug":"unspecified-ai-system","model_name":"Unspecified AI system","identity_class":"unknown","is_internal":false,"family_name":"Unspecified","family_slug":"unknown-unspecified","organization_name":"Unknown","organization_slug":"unknown"}],"sources":[{"id":"7eb5f4c4-1e48-4b46-98ba-cf2e0f976415","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.08.750024v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.08.750024v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-14T00:00:00.000Z","retrieved_at":"2026-09-15T03:00:54.418Z","title":"\n\nLucaCell: a sequence-centric foundation model for cross-species single-cell analysis \n\n","excerpt":"Single-cell foundation models have transformed transcriptomic analysis, yet most rely on fixed gene identifiers that limit transfer across species and data types. Here we present LucaCell, a sequence-centric foundation model that represents genes through pre-trained mRNA sequence embeddings rather than static gene annotations. Gene expression is discretized into bins and modeled with a Transformer encoder, enabling sequence-informed cell representation without a fixed gene-ID vocabulary. Pre-tra","content_hash":null,"source_reliability_class":"medium","is_primary_source":true,"is_independent":true,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-15T03:00:54.418Z"},{"id":"2710870a-8dec-4318-81ae-6902f0256a01","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.11.750964v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.11.750964v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-16T00:00:00.000Z","retrieved_at":"2026-09-17T07:00:55.431Z","title":"\n\nEstimand-aware and donor-aware triangulation of predefined gene-set signals across human tendon transcriptomic datasets \n\n","excerpt":"We analysed five predefined programmes across three human tendon transcriptomic contexts: acute exercise-associated fibroblast pseudobulks (E-MTAB-15400; 4 control and 4 exercise samples), mechanically stretched donor-derived cells (GSE150482; 3 normal and 3 tendinopathy donors), and paired lesional versus grossly normal-appearing tendon (GSE26051; 23 donor pairs). Inference was conducted at the sample, donor or donor-pair level. Non-stimulated GSE150482 single-cell RNA sequencing comprised only","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-17T07:00:55.431Z"},{"id":"6aec6fba-c59c-4f04-aa81-ef132e42fd78","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.11.750957v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.11.750957v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T21:01:00.119Z","title":"\n\nMapping Gene Expression to an Interpretable Semantic Space \n\n","excerpt":"Cell embeddings organize single-cell expression data, but their dimensions have no biological meaning, so clusters are interpreted afterward. We present MESIC (Mapping Expression to Semantic space with Interpretable Components), which builds the written knowledge about genes held in curated databases into the dimensions themselves. A biomedical language model converts each gene's summary into a semantic embedding. MESIC compresses these embeddings into a small number of components, each concentr","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T21:01:00.119Z"},{"id":"21272a92-c4c4-4205-9100-dc6a4d046ff4","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.15.751768v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.15.751768v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T22:01:01.781Z","title":"\n\nPretrained gene representations transfer mean expression more broadly than spatial patterns in virtual spatial transcriptomics \n\n","excerpt":"Models that combine tissue images with pretrained gene representations aim to predict spatial expression for genes not used to fit the downstream predictor. Yet success on held-out genes can reflect two capabilities: estimating a gene's mean expression across tissue locations and recovering its spatial variation. Across four cohorts spanning three human brain regions and HER2-positive breast cancer, we evaluated held-out genes in held-out individuals and separated these components. For spatial p","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T22:01:01.781Z"},{"id":"db349597-27ff-4efb-a545-c5573db38c03","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.17.752312v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.17.752312v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-23T00:00:00.000Z","retrieved_at":"2026-09-23T20:02:56.795Z","title":"\n\nMIRCID: Inferred Hub-miRNAs Drive Cross-Task Improvements in Drug Mechanistic Modeling \n\n","excerpt":"Drug mechanism-of-action (MoA) modeling commonly relies on perturbational transcriptomes, but matched microRNA (miRNA) measurements are often unavailable. Inferred regulatory features offer a scalable way to reuse these data. Here, we present MIRCID, a framework comparing gene expression with inferred transcription factor (TF) activity and miRNA expression across pathway classification and similarity-based MoA retrieval. HubmiRNet infers 414 pan-cancer hub miRNAs (HubmiRs) from 977 L1000 landmar","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-23T20:02:56.795Z"},{"id":"e3ae3f3e-8fb5-476e-9b8e-3d0282fb5e0c","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.17.752497v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.17.752497v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-24T00:00:00.000Z","retrieved_at":"2026-09-24T23:00:55.911Z","title":"\n\nISP Platform powered by Geneformer: Framework for Cross-Species, Sequential, and Multi-Gene In Silico Perturbation Screens with Application to iPS Cell State Transitions \n\n","excerpt":"In silico perturbation (ISP) enables virtual genetic screens, prioritizing candidate regulators before costly wet-lab experiments. However, applying foundation models requires programming expertise. Here, we present the ISP Platform, that automates ISP powered by Geneformer without code execution. The platform enables cross-species analyses by independently selecting input species and human or mouse Geneformer models. The platform also implements sequential multi-gene ISP, in which each perturba","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-24T23:00:55.911Z"},{"id":"df70eff8-3d31-4804-a5be-4472738b24cf","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.24.753701v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.24.753701v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-24T00:00:00.000Z","retrieved_at":"2026-09-24T23:00:56.183Z","title":"\n\nFrom single-sequence structure prediction to protein fitness landscape through a composable, epistasis-aware mutation atlas \n\n","excerpt":"Mapping the multi-mutant fitness landscape is vital to protein engineering, but is challenging due to the vast combinatorial sequence space awaiting exploration. A central difficulty lies in the accurate and efficient modeling of non-additive epistatic effects among individual mutations, which partially arise from the physical inter-residue interactions prescribed by the protein structure. Existing fitness predictors usually perform well on single mutants but become less powerful for higher-orde","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-24T23:00:56.183Z"},{"id":"cc4c1ae3-15d0-4aee-bd46-165221d7060b","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.18.752788v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.18.752788v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-24T00:00:00.000Z","retrieved_at":"2026-09-24T23:00:56.516Z","title":"\n\nASOFormer: a Transformer-based model for predicting antisense oligonucleotide efficacy to support therapeutic candidate prioritization \n\n","excerpt":"Antisense oligonucleotides (ASOs) represent a promising therapeutic modality for RNA-based disease mechanisms, but identifying high-efficacy candidates from large sequence spaces remains a major bottleneck in drug development. Here, we present ASOFormer, a novel Transformer-based neural network that predicts the inhibition efficiency of RNase H-mediated ASOs from sequence, predicted secondary structure, and chemical modification features. ASOFormer was trained on a knockdown efficacy dataset spa","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-24T23:00:56.516Z"},{"id":"e325fbc9-4b1c-455d-b9ad-5c63fae12a53","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.18.752645v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.18.752645v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-24T00:00:00.000Z","retrieved_at":"2026-09-24T23:00:56.760Z","title":"\n\nBoltz2-Notebook: An Interactive Google Colab Platform for Diffusion-Based Biomolecular Structure and Binding Affinity Prediction using the Boltz2 model. \n\n","excerpt":"Recent advances in deep learning-based structure prediction, including AlphaFold3 and the open-source Boltz model family, have extended biomolecular modeling to joint prediction of protein-ligand, protein-nucleic acid, and multi-chain complexes with binding-affinity estimation. Boltz-2 is among the most feature-complete of these open models, but its practical use requires a local CUDA-capable GPU, command-line execution, and manually authored YAML configuration files, limiting accessibility for ","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-24T23:00:56.760Z"},{"id":"be18af40-62c9-4e46-8ee2-2fab7c07b973","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.17.751511v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.17.751511v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-24T00:00:00.000Z","retrieved_at":"2026-09-24T23:00:56.991Z","title":"\n\nSurfGraphPro: Integrating Protein Language Models with Geometric Deep Learning On Coarse Protein Surfaces for Binding Site Prediction \n\n","excerpt":"Accurate and fast prediction of protein binding sites remains essential for understanding molecular interactions and facilitating protein engineering. Here, we present SurfGraphPro, a geometric deep learning approach that bridges protein language models with surface-based structural representations for a rapid binding interface prediction. Our method operates on coarse-graphed triangulated protein surfaces, utilizing solvent-excluded surface meshes downsampled into amino acid residue centered pa","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-24T23:00:56.991Z"},{"id":"4848e494-eb65-4ae8-b5e3-6dad4cacb400","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.24.754019v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.24.754019v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-24T00:00:00.000Z","retrieved_at":"2026-09-25T00:00:54.975Z","title":"\n\nDe novo design of small cyclic peptide inhibitors from non-canonical amino acids by cofolding-guided search \n\n","excerpt":"Cyclic peptides can engage targets beyond the reach of small molecules, but their discovery typically relies on libraries reused across targets, diversification around a known interaction motif, or computational methods limited to a handful of non-canonical building blocks. Here, we present nCycle-Forge, a simulated annealing framework that designs cyclic peptide binders by iteratively substituting building blocks and scoring each proposal by cofolding with the target. We designed, synthesized, ","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-25T00:00:54.975Z"},{"id":"d6a810e3-2528-4a33-9a7b-92e056a6f775","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.18.752572v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.18.752572v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-24T00:00:00.000Z","retrieved_at":"2026-09-25T00:00:55.355Z","title":"\n\nA Statistical Guideline for Analyzing Repeated-Measures Proteomics Data \n\n","excerpt":"Mass spectrometry-based proteomics increasingly supports study designs with repeated measurements, such as longitudinal sampling and spatial tissue profiling. These designs offer rich biological insights but introduce statistical challenges due to within-subject correlations and variance heterogeneity across time points or anatomical locations. Despite the growing use, best practices for analyzing repeated-measures proteomics data remain limited. We systematically compare commonly applied method","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-25T00:00:55.355Z"},{"id":"e8a56dd2-d398-43f8-89b6-64201d49ef04","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.16.750421v1?rss=1\n","canonical_url":"\nhttps://www.biorxiv.org/content/10.64898/2026.09.16.750421v1?rss=1\n","source_type":"preprint","publisher":"bioRxiv","author":null,"publication_date":"2026-09-24T00:00:00.000Z","retrieved_at":"2026-09-25T00:00:55.619Z","title":"\n\nAn interpretable open platform for sequence-based antibody developability prediction \n\n","excerpt":"Antibody developability is increasingly predictable from sequence, yet software and trained models are rarely made available. We present DELPHI, open software for training developability predictors from labelled antibody assay data, together with ready-to-run, retrainable models. DELPHI compares 25 language-model and classifier combinations under CDR H3-cluster cross-validation that reduces sequence-similarity leakage, measures how performance changes with labelled training-set size, and reports","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-25T00:00:55.619Z"}],"claims":[{"id":"40abc877-a3ab-4918-a81e-f38b3e724a8e","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","claim_text":"\n\nLucaCell: a sequence-centric foundation model for cross-species single-cell analysis \n\n","claim_type":"outcome","claim_status":"supported","confidence_score":null,"created_at":"2026-09-15T03:00:54.425Z","updated_at":"2026-09-15T03:00:54.425Z"}],"revisions":[{"id":"cfa00d4e-8731-436a-ace5-82c63d7ddab6","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.000000","new_score":"0.001000","previous_factors":{},"new_factors":{"evidence":0.1,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Auto-published from news ingest.","trigger_type":"news_ingest","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-15T03:00:54.431Z"},{"id":"e30ff2d2-b80a-4a57-8dd8-b94251264b13","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.001000","new_score":"0.002000","previous_factors":{"tier":1,"evidence":0.1,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.2,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["2710870a-8dec-4318-81ae-6902f0256a01"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-17T07:00:55.460Z"},{"id":"d8d3dd6b-82ca-4dae-97f1-39486a052558","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.002000","new_score":"0.003000","previous_factors":{"tier":1,"evidence":0.2,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.3,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["6aec6fba-c59c-4f04-aa81-ef132e42fd78"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T21:01:00.149Z"},{"id":"a4ed918a-5c69-4b78-9109-93c2562da4fd","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.003000","new_score":"0.004000","previous_factors":{"tier":1,"evidence":0.3,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.4,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["21272a92-c4c4-4205-9100-dc6a4d046ff4"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T22:01:01.811Z"},{"id":"f7105d0f-dca8-43f2-9d1e-43c223c33ef6","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.004000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.4,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["db349597-27ff-4efb-a545-c5573db38c03"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-23T20:02:56.865Z"},{"id":"d877a381-2dd6-4062-af85-44adeb07be47","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["e3ae3f3e-8fb5-476e-9b8e-3d0282fb5e0c"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-24T23:00:55.935Z"},{"id":"fc4fed7c-3a79-4879-a665-2b0bbf72a4e3","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["df70eff8-3d31-4804-a5be-4472738b24cf"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-24T23:00:56.201Z"},{"id":"b08f872e-3ac4-474a-a9a1-d053767ca6f2","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["cc4c1ae3-15d0-4aee-bd46-165221d7060b"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-24T23:00:56.531Z"},{"id":"3d2bc74f-9a00-4a92-89d2-e0cbae73173c","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["e325fbc9-4b1c-455d-b9ad-5c63fae12a53"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-24T23:00:56.775Z"},{"id":"f84cf646-b362-4a78-9e9a-300515c30522","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["be18af40-62c9-4e46-8ee2-2fab7c07b973"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-24T23:00:57.007Z"},{"id":"6fcd832d-f617-47d5-8586-419db8b47289","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["4848e494-eb65-4ae8-b5e3-6dad4cacb400"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-25T00:00:55.079Z"},{"id":"705972cb-a8ab-47da-b432-88b96ccbc240","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["d6a810e3-2528-4a33-9a7b-92e056a6f775"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-25T00:00:55.375Z"},{"id":"f0361339-cbb4-46b0-b6a6-a50843c0f0a0","event_id":"633d2ac4-5cfb-4893-ba54-8f30ef5c1213","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["e8a56dd2-d398-43f8-89b6-64201d49ef04"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-25T00:00:55.636Z"}],"secondary":[]},"methodology_version":"0.1"}