Biology subjects

Fu, Y.

Publications and source records attributed to Fu, Y..

At least 37 records · Page 2Linked to original sources

bioRxiv2026

Respiration-Deficient Cells Require Pyruvate Carboxylase to Suppress Asparagine Auxotrophy

Mutations in mitochondrial DNA (mtDNA) compromise ETC activity and impair oxidative phosphorylation. Since eukaryotic cells contain multiple copies of mtDNA, the resulting phenotype depends on the proportion of mutant mitochondrial genomes (the heteroplasmy level). Using isogenic cell lines carrying similar mtDNA deletions, a linear decline in cellular respiration was observed as mitochondrial DNA heteroplasmy increased. Despite this, cellular redox imbalance did not change until heteroplasmy exceeded 50%. As heteroplasmy increased past 70%, cells also exhibited an integrated stress response (ISR) and impaired translation was observed. These defects were reversed by either addition of asparagine or overexpression of pyruvate carboxylase (PC). The dependence on exogenous asparagine in other respiration-deficient cells was found to correlate inversely with the PC expression level. For example, patient-derived thyroid tumor cells, harboring high heteroplasmy for a Complex I mtDNA mutation and low levels of PC, exhibited asparagine auxotrophy, and L-asparaginase treatment suppressed tumor growth. Together, these findings demonstrate a role for mitochondrial pyruvate carboxylase in cellular asparagine synthesis under conditions of compromised respiratory activity.

cell biology↗

bioRxiv2026

Virtual-cell verification enables self-auditing AI discovery for immune rejuvenation

Artificial-intelligence agents propose drug-discovery hypotheses faster than experiments can test them, yet their conclusions are rarely verified, against the underlying biology, the predicted perturbation, or the agent's own scoring logic. We close this verification gap with an agentic framework built on three verifiers. First, PACE, a phenotype verifier, resolves immune aging into ten directionally scored, cell-type-resolved gene-set modules, selected for cross-cohort stability across four PBMC cohorts, and outperforms five established aging clocks in an independent in-house aging cohort of 434 elderly donors. Second, CellQ, a virtual-cell verifier built with multi-modal LLM, compresses each single-cell transcriptome into eight discrete tokens aligned to a language model's vocabulary through residual vector quantization; it attains state-of-the-art perturbation prediction and uniquely resolves the weak, module-level shifts that differential-expression recovery misses. Third, an Analyzer-Planner-Auditor agent verifies its own scoring logic: screening 110 compounds in primary human PBMCs, it found aged-down modules more reversible than aged-up modules and revised its objective from an equal-weight mean to a balance-constrained minimum, a self-correction that generalized to an independent 13-compound T-cell assay. By verifying its predictions and its own objective against experiment, the framework points beyond hypothesis-generating AI toward self-correcting AI scientists whose objectives could continuously evolve.

bioinformatics↗

bioRxiv2026

Three-dimensional vascular microenvironments uncover endothelial plasticity during TGF-β2-driven vascular remodeling

The tumor microenvironment plays a pivotal role in tumor development, harboring elements such as endothelial cells, immune cells, fibroblasts, and soluble factors such as transforming growth factor-{beta} (TGF-{beta}) family. TGF-{beta} family regulates cell development and promotes tumor invasion, metastasis, angiogenesis, and endothelial-to-mesenchymal transition (EndoMT). Here, we investigate the effects of TGF-{beta} signaling on vascular remodeling using a three-dimensional (3D) vascular network in a microfluidic device. Using both a co-culture (3D-Co) and simplified endothelial monoculture (3D-CM), we demonstrate that TGF-{beta} signaling reduces the quality and functionality of the vasculature by regressing them. In addition, we observed the upregulation of EndoMT-related markers in mRNA and protein expressions, suggesting the induction of EndoMT in 3D vascular networks. The increased vascular permeability stimulated by TGF-{beta}2 also supports the loss of endothelial identity in the 3D-Co. Transcriptomic analysis revealed the coordinated activation of pathways associated with cell migration and EndoMT, along with the suppression of cell cycle progression. A comparative analysis of two-dimensional (2D) and 3D cultures revealed a fundamentally distinct endothelial response to TGF-{beta}2 in the 3D context, including metabolic reprogramming. These findings demonstrate that the 3D microenvironment critically modulates endothelial responses to TGF-{beta} and enables the emergence of vascular phenotypes not captured in 2D systems. This study provides a more physiologically relevant platform to investigate endothelial dysfunction and vascular remodeling.

bioengineering↗

bioRxiv2026

Iterative co-creation of harmonized human and non-human primate cellular and structural ontologies and 3D common coordinate frameworks for the basal ganglia

A major goal of the BRAIN Initiative Cell Atlas Network (BICAN) is to create a suite of foundational reference cell atlases and associated standards for human and non-human primate brains. Central to this goal is the creation of cross-species harmonized cellular taxonomies and structural parcellations with formal ontologies that can be mapped into 3D reference frameworks bridging neuroimaging and cellular and histological resolutions. We describe here an iterative approach, focused initially on the basal ganglia, to co-create structural and cellular ontologies in human, macaque and marmoset brains, including a Harmonized Ontology of Mammalian Brain Anatomy (HOMBA), and to map and refine structural parcellations into neuroimaging-based common coordinate frameworks. These references provide the framework for documenting and mapping all experimental sampling in BICAN, allowing analyses of cellular and molecular variation as a function of topographic position, and enabling comparisons of cellular, molecular and neuroimaging-based functional variation within and between primate species. HighlightsO_LIA hierarchical Harmonized Ontology of Mammalian Brain Anatomy (HOMBA) covering 2348 structures C_LIO_LIHOMBA-annotated 3D common coordinate frameworks (CCFs) of the basal ganglia across species C_LIO_LIHistologically informed 3D parcellation/atlas of 280 human subcortical structures indexed by HOMBA C_LIO_LIMapping and integration of structural, cellular and functional data with HOMBA and CCFs C_LI

neuroscience↗

bioRxiv2026

Extracellular vesicles drive cross-genus mycovirus transmission and suppress two fungal diseases

Mycoviruses are typically transmitted vertically through fungal reproduction or horizontally via hyphal anastomosis, but identical viruses in phylogenetically divergent fungi hint at unknown inter-species transmission mechanisms. Here, we identify fungal extracellular vesicles (EVs) as mediators of cross-genus mycovirus transmission. Using the hypovirulent Botrytis cinerea strain IBc-374 (harboring 16 mycoviruses) as a donor, we show that up to 13 mycoviruses are horizontally transmitted to Sclerotinia sclerotiorum during dual culture or plant co-inoculation, even though these two fungi belong to different genera and are generally considered incapable of hyphal anastomosis. Electron microscopy reveals abundant vesicle structures in IBc-374 hyphae, and nanoparticle tracking analysis showed that the strain secretes >11-fold more EVs than a virus-free strain. RT-PCR detects genomic RNAs of 7 mycoviruses in purified EVs, and the full-length viral genome in EVs was further validated with BcHV5 as example by fluorescence in situ hybridization and RT-PCR. Incubation of protoplast-derived germlings of S. sclerotiorum or B. cinerea with IBc-374 EVs leads to infection by 4-6 donor mycoviruses, demonstrating EV-mediated cross-genus transmission. Injection of mycovirus-carrying EVs into tobacco leaves followed by fungal inoculation also transmits two hypoviruses to both species. Application of IBc-374 hyphal fragment suspension significantly reduces lesion sizes caused by both pathogens on plants, and rescued two pathogens carrying multiple mycoviruses exhibit hypovirulence and impaired growth. Our findings reveal EVs as a cell-free vector that bypasses vegetative incompatibility, providing a mechanistic basis for cross-species viral spread and opening avenues for EV-based virus cocktails to control multiple fungal diseases. IN BRIEFWu et al. discover that fungal extracellular vesicles (EVs) can package and transmit multiple mycoviruses across genera from Botrytis cinerea to Sclerotinia sclerotiorum. EV-mediated delivery overcomes vegetative incompatibility barriers and reduces disease caused by both fungal pathogens, offering a cell-free strategy for mycovirus-based biological control.

microbiology↗

bioRxiv2026

Hydrogen sulfide dynamically upregulates copper uptake and localization

The reactivity of copper, an essential micronutrient that undergoes facile cycling between Cu1+ and Cu2+ redox states, is carefully controlled within the confines of protein binding sites, and by sequestration in storage vesicles, or harnessed to kill pathogens by active pumping of Cu1+ into phagosomes. We have discovered that hydrogen sulfide, a signaling metabolite generated in copious quantities at the host-microbiome interface, upregulates Cu accumulation in diffusely dispersed puncta across the cell, as visualized by X-ray fluorescence microscopy. The Cu is predominantly in the Cu2+ state with oxygen/nitrogen ligands. Cu import occurs via the non- canonical ZNT1 transporter, while export, following sulfide withdrawal, is ATP7A-dependent. Cu accumulates at the apices of colon crypts in a mouse model of elevated sulfide exposure due to SQOR deficiency in the intestinal epithelium, establishing in vivo relevance. Our study reveals that sulfide is a dynamic regulator of the Cu pool, stimulating Cu2+ influx into highly concentrated puncta.

cell biology↗

bioRxiv2026

Evolutionary dynamics of sex chromosomes and candidate sex-determining genes across the beetle phylogeny

The evolutionary dynamics of beetle sex chromosomes remain poorly understood. Using synteny analyses of 163 chromosome-level genome assemblies, we examine sex chromosome evolution across the beetle phylogeny. Beetles retain a highly conserved ancestral X chromosome region, even in superfamilies with extensive autosomal rearrangements and in species with enlarged neo-X chromosomes formed by fusions of autosomal segments. In contrast, Y chromosomes vary greatly in presence, size and gene content, driven by degeneration of ancestral regions and addition of neo-Y segments in some species. Neo-X and neo-Y chromosomes frequently occur and generally evolve independently, highlighting a flexible neo-sex chromosome dynamics. We identify putative sex-determining genes and propose a novel candidate, the highly conserved and sexually differentially expressed X-linked gene ELAV-like. Our results uncover conserved X elements, fragile Y chromosomes, and independent neo-X and neo-Y formation across 300 million years of beetle genome evolution.

genomics↗

bioRxiv2026

Data-adaptive three-dimensional deconvolution and evaluation for volumetric fluorescence microscopy

Optical fluorescence microscopy enables visualization of biological structures and dynamics. However, the intrinsic diffraction limit, especially axially, and depth-related scattering noise compromise the image resolution and fidelity. Computational 3D deconvolution is a promising approach for mitigating these issues, yet its execution is hindered by inaccurate and cumbersome theoretical modeling or experimental measurement of 3D point spread function (PSF), as well as ineffective 3D noise regularization. Furthermore, in the 3D super-resolution regime, there remains a lack of standardized tools for evaluating 3D super-resolution fidelity. Here, we present the 3D adaptive deconvolution and evaluation (3D-ADE) toolkit, which comprises 3D-Ada deconvolution with physics-oriented automatic 3D-PSF calibration, and 3D-SQUIRREL for 3D super-resolution quality assessment. It effectively resolves noise instability, eliminates the need for 3D-PSF calibration, and reliably assesses the fidelity of 3D resolution extension via deconvolution, physical, and deep-learning-based methods. Accessible via multiple software platforms, 3D-ADE enhances the versatility of 3D deconvolution and fills the gap in 3D super-resolution evaluation tools, and thereby advances volumetric fluorescence imaging applications.

bioengineering↗

bioRxiv2026

BoltzProt-1: Towards Efficient De Novo Binder Design with Good Developability

Designing binders against novel protein targets remains a central challenge in computational drug discovery. Here we introduce BoltzProt-1, a pipeline for generating protein binders, including nanobodies, with improved hit rates and favorable developability properties. At its core lie a refined iteration of BoltzGens generative model and a novel protein-protein interaction prediction model, BoltzPPI. Employing BoltzPPI instead of BoltzGens standard structure-prediction confidence metrics to rank nanobody (VHH) designs increases the confirmed-binder hit rate from 3.3% to 8.0% across 10 novel targets. Assessed on 10 additional targets used in prior literature, the BoltzProt-1 pipeline obtains nanobody screening hits for 7 of 10 targets, surpassing the 6 of 10 previously reported by Chai-2. Finally, evaluating the developability of BoltzProt-1-designed nanobodies in terms of stability, aggregation, purity, polyspecificity and hydrophobicity reveals that 58% of its confirmed binders pass every criterion, exceeding both BoltzGen (40%) and clinical-stage VHH controls (21%). O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=104 SRC="FIGDIR/small/733997v1_ufig1.gif" ALT="Figure 1"> View larger version (39K): org.highwire.dtl.DTLVardef@d86de1org.highwire.dtl.DTLVardef@115da7eorg.highwire.dtl.DTLVardef@1bbb4e9org.highwire.dtl.DTLVardef@626a68_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2026

From hotspot dependence to distributed robustness in resistance-aware lead optimization

Drug resistance remains a recurrent failure mode in targeted anticancer and antiviral therapy, and resistance evidence often enters only after compound selection. ResistAgent is an evidence-constrained framework that converts mutational liabilities into design-time objectives through site- and combo-aware resistance mapping, deterministic mechanism diagnosis and robust counter-design. In EGFR-Erlotinib and HIV-RT-Rilpivirine, the framework separated residue-level liabilities from observed HIV combination liabilities and linked prioritized mutations to anchor loss, pocket rearrangement, electrostatic shifts and contact redistribution. Same-budget paired searches showed that robust objectives changed lower-tail mutant-panel behavior and interaction-dependence profiles while prioritizing robustness over average-affinity behavior. Under predefined liability panels, robust-objective trajectories shifted support away from mutable hotspot contacts toward more distributed interaction networks. Supplementary physical summaries and ranking-first benchmarks support the validation envelope of this resistance-aware design strategy.

bioinformatics↗

bioRxiv2026

Chromosome-Specific Expansion and Diversification of the Thionin Gene Family in Barley

Thionins are cysteine-rich peptides involved in plant defense. However, their genomic organization, evolutionary expansion, and potential function in barley remain unclear. Here, we integrated reference genome, pan-genome, and pan-transcriptome resources to systematically characterize the thionin gene family in barley. Fifty-six thionin genes were identified in the reference genome Morex V3, displaying pronounced chromosomal clustering and high sequence conservation consistent with extensive tandem duplication. Promoter analysis of these genes revealed enrichment of cis-acting elements associated with stress- and hormone-related signaling pathways, suggesting a potential role for thionins in biotic stress responses, including aphid defense, as suggested by previous studies. Analysis of 20 barley genotypes revealed substantial copy number variation, particularly on chromosomes 6H and 7H, indicating dynamic population-level expansion. Sequence-based clustering grouped thionins into ten clusters and five singletons, with major clusters corresponding to specific chromosomes. Integration of pan-transcriptome data showed that transcriptional activity was largely confined to four major clusters. Aphid infestation of four genotypes featuring copy number variation in chr 6H thionin genes resulted in a strong induction of thionin gene expression, with more pronounced responses during poor-host interactions. Aphid-induced expression tended to increase with thionin gene number, however, this no correlation was observed regarding basal gene expression levels. Together, these findings indicate that the thionin gene family in barley has undergone species-specific expansion driven by tandem duplication and contributes to genotype-dependent aphid defense responses.

Plant Biology↗

bioRxiv2026

Physics-Driven Zero-Shot Reconstruction of Isotropic 3D Fluorescence Microscopy under Undersampled Acquisition

Three-dimensional (3D) imaging represents the development of next generation of fluorescence microscopy. However, routine axial down-sampling makes isotropic resolution unrealistic. Here, we propose DeepUI, a physical zero-shot framework designed to achieve isotropic 3D fluorescence images from a low axial sampling rate. DeepUI fully leverages the intrinsic characteristics of 3D images through physics-guided degradation, which incorporates spatial-frequency joint learning to generate a scaled optical transfer function, combined with noise degradation and an up-sampling branch. Typically requiring just 5 minutes for training and 0.5 minutes for high-throughput and fast prediction, we demonstrate the superior performance of DeepUI to get isotropic results, and the exclusivity to axial down-sampling conditions, even in more challenging conditions, including defocused background, noise, and resolution blur.

bioinformatics↗

bioRxiv2026

Ground Truth-Based Evaluation of False Discovery Rate and Statistical Power in DIA Proteomics

Data-independent acquisition (DIA) mass spectrometry enables rapid proteomic quantification, yet the reliability of statistical inference in DIA-based protein quantification remains incompletely understood. Here, we systematically evaluated missingness, false discovery rate (FDR), and statistical power, defined as true positive rate (i.e. sensitivity or recall), using technical replicates and a spike-in benchmark with known ground truth. Analysis of 18 HeLa replicates revealed persistent, abundance-dependent missingness. In the spike-in experiment with five replicates, human peptides were titrated against a stable yeast background, allowing fold changes (FCs) to be compared with expected values. Across comparisons with log2FCs ranging from 0.2 to 2.5, the nominal BH-FDR substantially underestimated the true FDR. For example, at a BH-FDR threshold of 0.05, the true FDR was [~]0.2. Statistical power was [~]40% for a log2FC of 0.2 and increased to nearly 100% for a log2FC of 2.5. Additional incorporation of FC thresholds improved the true FDR for large-FC comparisons, with slight loss of power, but markedly reduced sensitivity for small-FC comparisons. Together, these results indicate that nominal FDR does not necessarily reflect actual error rates in DIA proteomics and that DIA performance is influenced by protein abundance and expected fold changes. This study provides a framework for experimental design and data interpretation in DIA-based proteomic studies.

bioinformatics↗

bioRxiv2026

Trichome acylsugar defenses are equally effective against leaf-mining and free-living insects

Trichomes are epidermal projections, crucial components of plant defense systems. Whether these epidermal defenses are effective against leafminers, which infest the internal mesophyll layers, remains unclear. Using the devasting invasive tomato pest, Phthorimaea absoluta (tomato leafminer), and the trichome-rich wild tomato, Solanum pennellii, we demonstrate that trichome produced acylsugars effectively inhibit the growth of endophagous P. absoluta larvae feeding inside leaves. High spatial resolution leaf- and frass-metabolite analysis by MALDI-MSI and Stimulated Raman Scattering microscopy revealed that trichome-synthesized acylsugars are translocated into the mesophyll tissues when larvae infest leaves. We identified an ATP-binding cassette (ABC) transporter, SpABCB5, as the mediator of this translocation and as an essential component of defense against leafminers. These data reveal that epidermal trichomes not only defend against free-living (ectophagous) herbivores through direct contact or rapid deposition of chemical defenses onto leaf surfaces, but also defend against endophagous leafminers by rapidly depositing chemical defenses into internal tissues. By delivering mines against leaf miners, trichomes offer new avenues for breeding crops with resistance to leafminers.

plant biology↗

bioRxiv2026

Inhibition of hexokinase 2 undermines cartilage health and accelerates osteoarthritis

ObjectiveEnhanced glycolysis is a metabolic hallmark of chondrocytes in osteoarthritis (OA); however, the roles of the glycolytic rate-limiting enzyme hexokinase 2 (HK2) in cartilage remain poorly understood. MethodsPharmacological approach (3-bromopyruvate (3-BrPA) treatment) and mice model involving HK2 knockout in Col2a1-expressing chondrocytes are utilized to access the impact of HK2 blockage on cartilage ex vivo and in vivo. The in vivo effects of HK2 inhibition on OA progression were evaluated using a destabilization of the medial meniscus (DMM)-induced OA mouse model, through both intra-articular 3-BrPA administration and chondrocyte HK2 deletion. Additionally, we analyzed published single-cell RNA sequencing (scRNA-seq) datasets from human articular cartilage and integrated these with bulk RNA-seq data from HK2-deficient chondrocytes to characterize HK2 expression features across conditions. ResultsBoth pharmacological inhibition and genetic deletion of HK2 impair cartilage formation ex vivo. Bulk RNA-seq analysis and ex vivo studies demonstrated a promoted ossification-like process due to HK2 ablation in chondrocytes. Through pseudotime analysis of published single-cell RNA sequencing (scRNA-seq) datasets from human articular cartilages, we further identified that HK2 is differentially expressed across conditions, with a feature of a relatively high expression level at terminal stages of chondrocyte differentiation in the context of OA. We next confirmed HK2 deficiency in chondrocytes significantly exacerbated OA progression but having no impact on skeletal development in mice. ConclusionsHK2 plays a critical role in maintaining cartilage health, likely through the regulation of calcification, thereby highlighting the potential risks associated with targeting glycolytic enzymes as a therapeutic strategy for OA.

pathology↗

bioRxiv2026

dbGIST: An LLM-Assisted Multi-Omics Resource for Target Exploration and Cross-Dataset Validation in Gastrointestinal Stromal Tumors

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal neoplasms of the gastrointestinal tract, yet GIST-specific omics evidence remains scattered across small cohorts and is not represented as a dedicated disease project in major cancer genomics resources, limiting reproducible target exploration. Here, we present dbGIST (https://www.dbgist.com), a dedicated GIST-focused multi-omics resource built to make dispersed GIST evidence searchable, analyzable, and reusable. dbGIST harmonizes data from 37 centers and 1,991 samples, including pathologically verified in-house cohorts, across genomics, bulk transcriptomics, proteomics, phosphoproteomics, and single-cell transcriptomics, and couples these data with curated clinical annotations covering survival, mutation status, risk stratification, metastasis or recurrence, mitotic index, tumor site and size, and imatinib response. The platform supports cohort-level molecular-clinical association, survival, enrichment, immune-infiltration, drug-sensitivity, and single-cell analyses through interactive visualizations, downloadable source data, and public APIs for programmatic access to reusable analysis outputs and visualization-ready data. An optional LLM-assisted interface helps users navigate analyses and interpret outputs. Using MCM7 as a case study, dbGIST linked a resource-derived candidate to survival, risk features, metastatic or recurrent disease, imatinib-response phenotypes, proliferative cell states, and in vitro GIST-cell behavior. dbGIST therefore provides a traceable and interoperable resource for target exploration and precision oncology research in GIST.

cancer biology↗

bioRxiv2026

Somatic copy-number alteration signatures reveal sex-divergent aging trajectories accelerated in cancer

Somatic copy-number alterations (CNAs) accumulate with age and contribute to age-related pathologies, but their systematic characterization at single-cell resolution has been limited by the throughput-resolution trade-off in single-cell whole-genome sequencing. Here, we developed ultra-CNA, a high-resolution single-cell analysis pipeline that extends CNA detection to 10-kb bin resolution and jointly profiles copy-number and single-nucleotide variation (SNV). Re-analyzing the Tasc-WGS dataset (Liu et al., 2022; previously analyzed at 200-kb resolution) of 32,526 lymphocytes from 16 healthy donors aged 0.7 to 79 years, we constructed a multi-dimensional CNA spectrum stratified by chromosomal context, copy-number state, size, and clonality. Small (<1 Mb), rare, predominantly loss-type CNAs accumulated progressively and stochastically with age. Sex-chromosome loss showed divergent kinetics: chromosome X loss cells in females accumulated at +0.10 percentage points per year, versus +0.03 for chromosome Y loss cells in males. Sex chromosome loss also had specific consequences for autosomal SNV burden: in younger donors, loss cells carried fewer autosomal SNVs than non-loss cells, whereas in older donors (>30 years), loss cells exceeded non-loss cells in both sexes. Female X-loss cells additionally exhibited elevated 45S rDNA copy number, supporting biologically distinct consequences of X loss and LOY. Clock-like SBS1 and SBS5 mutational signatures co-accumulated with age across both sexes. Applying KL-divergence non-negative matrix factorization to the channelized CNA spectra, we constructed an aging clock validated by leave-one-sample-out cross-validation. Applied to a matched esophageal cohort, the clock detected accelerated aging from normal squamous epithelium through Barretts esophagus to esophageal adenocarcinoma, with cancer-associated spectra additionally enriched for large, highly clonal events. Ultra-CNA thus provides a scalable framework for quantifying somatic genomic aging from blood and for detecting accelerated aging in cancer.

genomics↗

bioRxiv2026

Towards autonomous biology: Compiler-Verified Protocols as a Foundation for Real World AI Execution

Artificial intelligence has advanced from analyzing experimental data to autonomously generating hypotheses, designing experiments, and coordinating closed loop discovery. Yet the translation from computational reasoning to physical execution remains bottlenecked by the experimental protocol, which in biology still relies on ambiguous natural-language descriptions: a medium other engineering disciplines abandoned decades ago in favor of compiler verified specification languages. This deficit fragments reproducibility along three axes: protocol accuracy, pre execution verification, and cross platform portability. Existing formalisms address only subsets of these challenges, trading expressiveness for rigor, portability for standardization, or usability for provenance. Here we introduce the Biology Protocol Language (BPL), a domain specific language with a biology-native type system in which every quantity carries physical units, every reagent declares its physical form, and every container maintains compiler-tracked state, so that implicit assumptions must be stated explicitly and physically impossible operations are rejected at compile time. We further develop BPL-COGEN, a pipeline that couples a fine tuned 30 billion parameter language model with the deterministic compiler in a closed generate validate repair loop, iteratively correcting the translation from natural language SOPs to BPL through compiler diagnostics until all physical, dimensional, and state constraints are satisfied. On a benchmark of 300 published Nature Protocols papers, BPL COGEN achieved an overall fidelity score of 95.1 against the source protocols as ground truth. Wet-lab experiment and cross-platform validation in GFP expression library construction and HPLC to UHPLC method translation confirmed that a single BPL source yielded reproducible execution across manual and liquid handler assisted contexts. The results established a novel pipeline that generates compiler-verified protocols, which is an essential prerequisite for physically embodied AI in biology.

synthetic biology↗