Biology subjects

Xu, L.

Publications and source records attributed to Xu, L..

At least 127 records · Page 7Linked to original sources

bioRxiv2025

SKiM-GPT: Combining Biomedical Literature-Based Discovery with Large Language Model Hypothesis Evaluation

BackgroundGenerating and testing hypotheses is a critical aspect of biomedical science. Typically, researchers generate hypotheses by carefully analyzing available information and making logical connections, which are then tested. The accelerating growth of biomedical literature makes it increasingly difficult to keep pace with connections between biological entities emerging across biomedical research. Recently developed automated means of generating hypotheses can generate many more hypotheses than can be easily tested. One such approach involves literature based discovery (LBD) systems such as Serial KinderMiner (SKiM), which surfaces putative A B C links derived from term co occurrence. However, LBD systems leave three critical gaps: (i) they find statistical associations, not biological relationships; (ii) they can produce false positive leads; and (iii) they do not assess agreement with a hypothesis in question. As a result, LBD search results often require costly manual curation to be of practical utility to the researcher. Large language models (LLMs) have the potential to automate much of this curation step, but standalone LLMs are hampered by hallucinations, lack of transparency in information sources, and the inability to reference data not included in the training corpus. ResultsWe introduce SKiM-GPT, a retrieval-augmented generation (RAG) system that combines SKiMs co-occurrence search and retrieval with frontier LLMs to evaluate user- defined hypotheses. For every chosen A-B-C SKiM hit, SKiM-GPT retrieves appropriate PubMed abstract texts, filters out irrelevant abstracts with a fine-tuned relevance model, and prompts an LLM to evaluate the users hypothesis, given the relevant abstracts. Importantly, the SKiM-GPT system is transparent and human-verifiable: it displays the retrieved abstracts, the hypothesis score, and a justification for the score grounded in the texts and written in natural language. On a benchmark consisting of 14 disease-gene-drug hypotheses, SKiM-GPT achieves strong ordinal agreement with four expert biologists (Cohens {kappa} = 0.84), demonstrating its ability to replicate expert judgment. ConclusionsSKiM-GPT is open-source (https://github.com/stewart-lab/skimgpt) and available through a web interface (https://skim.morgridge.org), enabling both wet-lab and computational researchers to systematically and efficiently evaluate biomedical hypotheses at scale.

bioinformatics↗

bioRxiv2025

Density dependence promotes species coexistence and provides a unifying explanation for distinct productivity-diversity relationships

Understanding diversity patterns in complex communities, such as microbial consortia, requires a mechanistic framework appropriate for many species. Negative density dependence is often utilized in complex ecosystem models, typically as a density-dependent mortality term for a population, but its full impact on community structure remains unclear. Here we use mechanistic population models of resource consumption to examine the effects of negative density dependence and develop a tractable framework for understanding diversity patterns in complex systems. To provide mechanistic grounding, we quantify how density-dependent mortality expands coexistence zones along resource gradients in simple communities using graphical analysis. We then derive an analytical, ecologically insightful formula predicting species abundances in subsets (guilds) of complex communities, in which many species share a resource or predator. Finally, we use the formula to explain how distinct relationships between productivity and diversity emerge from the resulting mechanistic framework, providing insights into previously unreconciled observed patterns.

ecology↗

bioRxiv2025

Improvements to Casanovo, a deep learning de novo peptidesequencer

Casanovo is a state-of-the-art deep learning model for de novo peptide sequencing from mass spectrometry proteomics data. Here we report on a series of enhancements to Casanovo, aimed at improving the interpretability of the scores assigned to predicted peptides, generalizing the software for use in database search, speeding up training and prediction runtimes, and providing workflows and visualization tools to facilitate adoption of Casanovo and interpretation of its results. Our goal is to make Casanovo accurate and easy to use for applications such as metaproteomics, antibody sequencing, immunopeptidomics, and discovery of novel peptide sequences in standard proteomics analyses. Casanovo is available as open source at https://github.com/Noble-Lab/casanovo.

molecular biology↗

bioRxiv2025

A Cattle BodyMap of Transcriptome across 52 Tissues and 3 Developmental Stages Reveals New Genetic Insights into Beef Production Traits

Cattle are important livestock that provide essential meat and milk resources. However, a comprehensive analysis of gene expression, alternative splicing (AS), and RNA editing across various organs and developmental stages in cattle has not been reported. This study aims to create a comprehensive transcriptomic BodyMap across various tissues and developmental stages, integrating this information into genomic predictions of beef production traits. We created a comprehensive transcriptomic BodyMap using 400 samples collected from 52 organs of newborn, young, and adult cattle, and estimated their contributions to genetic variance and genomic predictions for 23 beef production traits in 1476 beef cattle. We cataloged the expression of 25,530 annotated genes, 28,533 novel long non-coding RNAs (lncRNAs), 215,754 AS events, and 3,093,058 A-to-I RNA-editing sites. Integrating transcriptome BodyMap with 23 beef production traits, we found lncRNAs influenced traits like rib-eye area and carcass length, while RNA editing associated with chunk roll weight and daily gain. We observed trait-relevant tissues between different stages, including the differential expressed genes of cerebellum, longissimus muscle, and testis between newborn and adult stages are more relevant to beef production traits. The tissue-specific genes and development-associated genes in several tissues could improve the reliability of genomic prediction in beef production traits. We developed Cattle BodyMap Transcriptome Database (https://cattlegenomics.online/cattle_bodymap) to retrieve, analyze, and visualize gene expression, lncRNA, splicing and RNA editing data. Our results demonstrated the potential of using transcriptome data as a valuable resource for genomic selection and breeding programs in beef cattle. Additionally, our transcriptome BodyMap serves as a valuable resource for biological interpretation, functional validation, and genomic improvement in livestock.

genomics↗

bioRxiv2025

Complexes between Netrin G Ligands and Chiral Nanoparticles Promote Axons Regeneration under Near-Infrared Illumination

Chiral nanoparticles combine functionalities of inorganic materials and large biomacromolecules enabling stimulation of neurons. However, multiple challenges remain in their use for central nervous system including identifying suitable cell signaling pathways and traversing the blood-brain barrier. In this study, we show that glutathione-coated Ni(OH)2 nanoparticles form nanoscale complexes with netrin-G1 ligands (NGL-1), critical for neuronal regeneration. NGL-1 features a semispherical pocket with a 2.5 nm radius of curvature, fitting well with nanoparticles sized 3 {+/-} 1.2 nm. D-NPs with D-glutathione surface ligands activate D-glutathione receptors on epithelial cells, facilitating their transport into the brain. When illuminated with 980 nm light, the nanoparticle-protein complex stimulates axon regeneration through localized IGF-1 production. This approach successfully regenerated (a) hippocampal neurons in Alzheimers disease mice and (b) dorsal root ganglion neurons in spinal cord-injured mice. The nanoparticles were thoroughly tested for safety and excreted intact. The local, rather than systemic, IGF-1 upregulation minimizes its side effects.

neuroscience↗

bioRxiv2025

Host Microenvironment Reprogramming by Saccharides Overcomes Lung Barriers for mRNA Therapeutics

AbstractOvercoming biological barriers remains the paramount challenge for pulmonary mRNA therapeutics. Conventional approaches focus exclusively on passively optimizing formulation quality without controlling dynamic host barriers. Here, we pioneer a host-centric strategy by leveraging sugar that actively reprograms the airway microenvironment to boost IVT-mRNA transfection. Utilizing machine learning-accelerated screening of a chemically diverse saccharide library, we identify D-glucose as the best-performing candidate. Glucose assisted-delivery within lipid nanoparticles (Glu-LNP) achieves robust, lung-specific protein expression (up to 131.21-fold increase) across diverse preclinical models with reduced inflammation. In lung carcinoma models, Glu-LNP-encapsulated IL-12 mRNA reduced tumor burden by approximately 59.12% and improved survival by 2.5-fold compared to the LNP group. Mechanistically, glucose orchestrates a dual-pathway cascade: metabolic reprogramming via the Warburg effect elevates ATP, fueling endocytosis and translation; ATP further activates the P2Y2-IP3 signaling axis that triggers Ca2+ release and subsequent CLCA1/TMEM16A-dependent chloride/bicarbonate efflux, which remodels mucus barriers and enhances nanoparticle penetration. This bioenergetic and mucolytic host intervention strategy presents a broadly applicable paradigm to transcend delivery limitations for respiratory mRNA therapeutics.

bioengineering↗

bioRxiv2025

Methyl gallate Attenuates Post-Stroke Emotional and Cognitive Symptoms by Promoting Hippocampal Neurogenesis via PI3K/GSK3 and AMPK Signaling

Restoring hippocampal neurogenesis is an effective strategy for post-stroke recovery. Methyl gallate (MET) exhibits neuroprotective properties. However, the effect of MET in improving brain functional recovery in the post-stroke depression (PSD) model and its underlying mechanism remains unknown. Single-cell data analysis showed that the cell types and molecular characteristics of PSD are similar to those of primary depression but exhibit weaker synaptic plasticity and stronger inflammatory signals. In addition, molecular docking studies revealed that MET exhibits a significant binding capacity with AMPK/GSK3, suggesting that MET mediates the neuroprotective effects of both. In this study, we created a post-stroke depression (PSD) model by performing physical restraint after ischemia and tested the treatment effects of MET. We observed that MET significantly attenuated PSD-induced depressive-/anxiety-behaviors associated with a reduction of stress hormone corticosterone and ACTH levels. Morris water maze and recognition task results indicate that MET can also alleviate cognitive impairments in the PSD model. In the hippocampus of the PSD model, MET improved the proliferation and differentiation of neural stem/progenitor cells (NSPCs). MET treatment significantly enhanced the activity of AMPK and decreased the activity of GSK3{beta}. Furthermore, in primary neural progenitors under hypoxia, both the PI3K inhibitor LY294002 and the AMPK inhibitor compound C blocked the effects of MET to promote neural development. Animal experiments also confirmed that LY294002/compound C treatment could reduce the effects of MET in antidepressant behaviours. Taken together, our results indicate that PI3K, as well as AMPK-mediated adult neurogenesis, is restored by MET to improve brain functions in the PSD model.

neuroscience↗

bioRxiv2025

The RXLR-EER Motif Determines an Unconventional Secretion Pathway Associated with Extracellular Vesicle Production

Phytophthora infestans, the cause of potato late blight disease, delivers a suite of RXLR effectors into host plant cells to subvert immunity, whereas apoplastic effectors act extracellularly. Although the RXLR-EER motif in these effectors is critical for host translocation and is cleaved prior to secretion, the relevance of this processing is poorly understood. Prior evidence suggests RXLR effectors utilize a distinct, unconventional secretion pathway, raising the question of whether the RXLR-EER motif influences selection of the secretion route. Here, we combined genetic, molecular and cell biology approaches to investigate the secretion pathway of RXLR effectors. Confocal microscopy revealed that RXLR and apoplastic effectors localize to distinct vesicular compartments in cultured hyphae. Moreover, fusing the ER retention signal KDEL to RXLR effectors did not impair their secretion, in contrast to apoplastic effectors, which were retained in the endomembrane system, indicating that RXLR effectors bypass the canonical ER-to-Golgi pathway. Importantly, RXLR effectors associate with extracellular vesicles (EVs), whereas RXLR-EER motif mutants show reduced EV association and are rerouted through the ER-to-Golgi secretion pathway. These findings demonstrate that the RXLR-EER motif governs effector sorting into an unconventional, EV-linked secretion route. This study sheds light on the molecular basis of effector trafficking in P. infestans and underscores the potential role of EVs in delivering virulence factors during host colonization.

microbiology↗

bioRxiv2025

Almost Free Enhancement of Multi-Population PRS: From Data-Fission to Pseudo-GWAS Subsampling

Many multi-population polygenic risk score (PRS) methods have been proposed to improve prediction accuracy in underrepresented populations; however, no single method outperforms other methods across all data scenarios. Although integrating PRS results across multiple methods and populations may lead to more accurate predictions, this approach may be limited by the availability of individual-level tuning data to calculate combination weights. In this manuscript, we introduce MIXPRS, a robust PRS integration framework based on data fission principles, to effectively combine multiple multi-population PRS methods using only genome-wide association study (GWAS) summary statistics from multiple populations. Specifically, MIXPRS employs SNP pruning to mitigate linkage disequilibrium (LD) mismatch between the training GWAS summary statistics and LD reference panels, and utilizes non-negative least squares regression to robustly estimate PRS combination weights. Extensive simulations and real-data analyses involving 22 continuous traits and four binary traits across five populations from the UK Biobank and All of Us datasets demonstrate that MIXPRS consistently outperforms the existing methods in prediction accuracy. Because MIXPRS relies solely on GWAS summary statistics, it enjoys broad accessibility, robustness, and generalizability for underrepresented populations.

genetics↗

bioRxiv2025

Rapid acquisition of HIV-1 neutralization breadth in a rhesus V2 apex germline antibody mouse model after a single bolus immunization

Current vaccine strategies to elicit broadly neutralizing antibodies (bnAbs) against HIV-1 generally propose complex, multi-boost immunization regimens. In rhesus macaques, SHIV infection has been observed to rapidly drive the development of some classes of bnAbs that share structural similarities with those in humans. Here, we generated a knockin mouse model with B cells bearing the unmutated common ancestor (UCA) of the V2 apex-targeted bnAb lineage, V033-a. A single immunization of mice with a germline-targeting native-like trimer was sufficient to recapitulate the ontogeny of the mature rhesus bnAb in knockin mice--including rare, disfavored somatic mutations--leading to the induction of antibodies that exhibited potent neutralization against both autologous and heterologous tier 2 viruses. A boost with Env escape mutant trimers further improved breadth and potency, and cryo-EM structure revealed the structural basis for heterologous neutralization breadth. Non-human primate and mouse models can thus combine with structure to serve as a platform for identifying and confirming immunogens that streamline HIV-vaccination regimens.

immunology↗

bioRxiv2025

Genetically Programmable Adhesive Protein Hydrogels for Sealing Perforating Corneal Trauma

The cornea, situated at the forefront of the eyeball, is often subjected to varying degrees of mechanical laceration injuries due to various traumatic incidents. Timely and precise suturing of corneal wounds to prevent the leakage of intraocular contents is a significant clinical challenge. Although some studies have attempted to use hydrogels to accurately cover corneal injuries, most of these hydrogels are only suitable for adhering to focal stromal defects that do not involve the corneal stroma. Here, a fully protein-based hydrogel, based on SpyTag-SpyCatcher chemistry, was investigated for repairing corneal perforating injury models in rats and rabbits. The hydrogel formation relies on covalent bonding between SpyTag (A) and SpyCatcher (B) under physiologically mild conditions, eliminating requirements for organic solvents, elevated temperatures, or photoactivation Notably, integration of recombinant mussel foot protein 3 (Mfp3) endowed the hydrogels with exceptional water-resistant adhesiveness, while maintaining their original mechanical properties and biocompatibility. This study demonstrates the great potential of Mfp hydrogels in the timely and precise repair of corneal perforating injuries in complex physicochemical environments. TeaserBioinspired protein hydrogel rapidly seals corneal perforations, minimizing scarring and inflammation without sutures in animal models.

bioengineering↗

bioRxiv2025

Direct observation of single intrinsically disordered proteins in solution

Intrinsically disordered proteins (IDPs) and intrinsically disordered regions (IDRs) in structured proteins are integral to many biological processes including neurotransmitter regulation, microtubule regulation, and transcription. IDP/IDRs are heterogenous, existing in a conformational ensemble of various interconnected states without a definitive tertiary structure. The high dynamicity of IDPs/IDRs limits ensemble protein characterisation techniques from capturing their properties, and measurements at the single-molecule level are hampered by the necessity to label the protein or modify its microenvironment, affecting their biophysics. Consequently, our understanding of IDPs/IDRs is limited, translating to a lack of knowledge of their roles in related diseases including Alzheimers disease, Parkinsons disease, and various cancers. This work presents the first experimental observation of unmodified IDP/IDR conformational dynamics in vitro, at the single-molecule level in real time, achieved by trapping individual IDPs/IDRs in a nanoscale volume using nanoaperture optical tweezers. Our results reveal that IDPs/IDRs exhibit significantly larger conformational variations in solution compared to globular proteins of similar size, as expected. We demonstrate that phosphorylation of native tau-441 by glycogen synthase kinase 3-beta (GSK3{beta}-tau) induces compaction and reduced conformational dynamics. We further observed a disorder-to-order transition during binding of an IDR, the N-terminal region of Src-associated protein in mitosis of 68 kDa (Sam68), to G8.5 RNA. The capability of nanoaperture optical tweezers to monitor the dynamic behaviours of single, unmodified IDPs/IDRs provides a powerful approach to advance our understanding of their elusive behaviours and further decode their roles in associated diseases.

biophysics↗

bioRxiv2025

BrainVAE: Exploring the role of white matter BOLD in preclinical Alzheimer disease classification

INTRODUCTIONLike gray matter (GM), white matter (WM) BOLD functional signals change in preclinical AD. However, the potential of WM BOLD for identifying preclinical AD remains underexplored. METHODSWe developed BrainVAE, a transformer-based variational autoencoder with interpretability, to classify preclinical AD and normal controls using resting-state fMRI data. We benchmarked BrainVAE against nine alternative models under three input configurations: WM-only, GM-only, and combined WM+GM. Interpretability analysis was also performed to investigate each brain regions contribution to the classification task. RESULTSBrainVAE outperformed other models and performed well (accuracy = 83.42%, F1-score = 91.62%, AUC = 64.50%) using the combined input compared to WM-only and GM-only. Specific WM bundles--including corpus callosum, fornix, and corticospinal tract--were among the most influential features contributing to the classification. DISCUSSIONIncorporating WM BOLD signals improves the distinction of preclinical AD from controls, underscoring the potential role of WM BOLD features for detecting early-stage AD. HighlightsO_LIBrainVAE integrates white and gray matter BOLD signals for classification of pre-AD and controls. C_LIO_LIBrainVAE achieves high accuracy (83.42%) and F1-score (91.62%) in identifying pre-AD. C_LIO_LIModels using combined WM+GM inputs outperform those using WM-only or GM-only inputs. C_LIO_LIWM regions, such as corpus callosum and fornix, contribute significantly to model predictions. C_LIO_LIResults suggest WM BOLD signals are informative markers for early AD detection. C_LI

neuroscience↗

bioRxiv2025

Quantitative RNA pseudouridine landscape reveals dynamic modification patterns and evolutionary conservation across bacterial species

Pseudouridine ({Psi}) modifications are the most abundant RNA modifications; however, their distribution and functional significance in bacteria remain largely unexplored compared to eukaryotic systems. In this study, we present the first transcriptome-wide and quantitative mapping of {Psi} modifications across five diverse bacterial species (Bacillus cereus, Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Pseudomonas syringae) at single-base resolution, utilizing the optimized BID-seq method for bacterial RNA. Our analysis revealed growth phase-dependent dynamics of pseudouridylation in bacterial tRNA and mRNA, particularly in genes enriched in core metabolic pathways. Comparative analysis demonstrated evolutionarily conserved features of {Psi} modifications, such as dominant motif contexts, {Psi} clustering within operons, etc. Functional analysis indicated {Psi} modifications influence bacterial mRNA stability, translation, and interactions with specific RNA-binding proteins (RBPs) in response to changing cellular demands during growth phase transitions. The integrated computational analysis on local RNA architecture was conducted to elucidate the structure-dependent {Psi} modifications in bacterial RNA. Furthermore, we developed an integrated deep learning framework, combining Transformer-GNN-based neural networks (pseU_NN) to capture both RNA sequence and structural features for effective prediction of {Psi}-modified sites. Overall, our study provides valuable insights into the landscapes of bacterial RNA {Psi} modifications and establishes a foundation for future mechanistic investigations into the functions of {Psi} in bacterial RNA regulation.

microbiology↗

bioRxiv2025

Structural insight into the transcription activation mechanism of the phage Mor/C family activators

Bacteriophage Mu, a temperate phage that infects E. coli K-12 and other enteric bacteria, precisely controls its replication cycle through hijacking host RNA polymerase (RNAP) by the middle operon regulator Mor and the late gene transcription activator C. Though a dimeric arrangement and significant conformational changes are proposed for the distinct Mor/C family activators, the underlying transcription activation mechanism remains unclear. In this study, we present two cryo-EM structures of the transcription activation complex (Mor-TAC and C-TAC) with phage Mu middle and late gene promoters, respectively. Remarkably, the Mor/C activators bind to promoter DNA as a centrosymmetric tetramer rather than as the proposed dimer, concurrently stabilizing by the N-terminal dimerization domains and C-termini. The C-terminal DNA binding domains and two anti-{beta}-strands simultaneously interact with two adjacent DNA major grooves. The activators also engage a variety of interactions with the conserved domains (CTD, {sigma}70R4, and {beta} FTH) of RNAP, providing evidences for a recruitment mechanism. In addition, single-molecule FRET assays show that C significantly enhances RPitc formation, suggesting a different multi-step activation mechanism for C. Collectively, these findings reveal the unique transcription activation mechanism of tetrameric Mor/C family activators, unraveling a novel mode of phage hijacking and bacterial transcription regulation.

biochemistry↗

bioRxiv2025

Robust Bayesian Integrative Modeling of Single Cell and Spatially Resolved Transcriptomics Data

A recent technology breakthrough in spatially resolved transcriptomics (SRT) has enabled comprehensive molecular characterization at the cellular level while preserving spatial information. However, many SRT technologies (e.g., spatial transcriptomics) cannot achieve single-cell resolution. Instead, they measure the average gene expression of a mixture of cells within a spot arranged on a lattice. Here, we introduce iMOSCATO, a fully Bayesian model that integrates single-cell RNA sequencing (scRNA-seq) data and SRT data to simultaneously decompose cell-type compositions of spots and identify the underlying spatial domains. We incorporate the lattice structure by employing a Markov random field prior, improving the accuracy of both cell-type deconvolution and spatial domain detection. Moreover, we use a zero-inflated Dirichlet distribution to capture cell-type sparsity. Finally, iMOSCATO shows competitive performance in accuracy compared to existing methods in both simulation studies and two real data applications.

bioinformatics↗

bioRxiv2025

Personalized Brain Morphometric Feature as a Transdiagnostic Predictor of Psychopathology: Insights from Dual Systems Models

BackgroundAdolescents are particularly vulnerable to developing psychopathological symptoms, yet neurobiological markers that can identify these vulnerabilities in a personalized and interpretable manner remain limited. Dual Systems Models suggest that this vulnerability may result from asynchronous development of neural systems subserving cognitive-control and socioemotional functions. Given that rigorous empirical evidence is sparse, this study aimed to quantify developmental imbalances and evaluate their predictive value for psychopathology. MethodsBased on Dual Systems Models, the dorsolateral prefrontal cortex (DLPFC) and ventral striatum (VS) were selected as key regions for the two brain systems, respectively. Using longitudinal data from the Adolescent Brain Cognitive Development (ABCD) study (baseline: n=11,238, ages 9.92{+/-}0.625 years; 2-year follow-up: n=7,870, ages 12.2{+/-}0.652 years; 4-year follow-up: n=2972, ages 14.1{+/-}0.693 years), we derived a personalized imbalance score based on brain morphometric features, including surface area, thickness and gray-matter volume. Nested linear mixed models were employed to assess associations between the imbalance scores and psychopathology. Additionally, generalized additive models were used to capture developmental trajectories of the imbalance scores and explore potential non-linear associations with psychopathology. To examine reproducibility, data from the Lifespan Human Connectome Project in Development study (HCP-D, N=652, ages 8-21 years) were interrogated. ResultsThe imbalance score, quantified as the difference between VS volume and DLPFC surface area, exhibited strong reliability and validity. During adolescence, the score exhibited a declining trend, with decreasing growth rates and variability between individuals. The score was positively associated with externalizing symptoms and showed a U-shaped relationship with internalizing symptoms. These findings were replicated in the HCP-D sample. ConclusionsThe novel neuroanatomical imbalance score empirically supports Dual Systems Models and may function as a transdiagnostic marker for internalizing and externalizing psychopathology. These findings enhance understanding of the neurodevelopmental mechanisms underlying adolescent psychopathology and offer potential implications for precision prevention and intervention strategies.

developmental biology↗

bioRxiv2025

STM is required for fate establishment of productive shoot progenitor cells in Arabidopsis tissue culture

De novo shoot regeneration via the two-step tissue culture method is widely used for vegetative reproduction of plants and for gene transformation and editing. A key step in tissue culture is the fate transition from shoot progenitor cells (SPCs) to productive SPCs (prSPCs), which are capable of differentiating into a shoot. In this study, we identified the class I KNOTTED1-like homeobox (KNOX) transcription factor gene SHOOT-MERISTEMLESS (STM) as the key controller of, and marker gene for, the fate transition from SPCs into prSPCs, indicating that STM might serve as a candidate of molecular tools to improve de novo shoot regeneration in the agricultural applications.

plant biology↗