Browse Articles
Discover research articles across all indexed journals
Changes in El Niño–Southern Oscillation and global frequency entrainment
Automated High-Throughput Virtual Screening of Catalysts via Templated Organic Reaction Pathway Construction: A Case Study on Suzuki–Miyaura Coupling Reaction
A practical and safe alternative method for skeletal cleaning for museum specimens using superworms (Zophobas morio)
Clean and undamaged skeletons that maintain natural anatomical shape are essential for anatomical collections and natural history museums. Conventional methods, such as maceration, chemical treatments, or dermestid beetle colonies, although commonly used, often require long processing times, pose biohazard risks, and may damage delicate bones. This study explores the use of superworms ( Zophobas morio ) as a biological alternative for skeletal cleaning. Controlled cleaning trials were conducted using specimens from various vertebrate groups and a range of size classes, categorized as small, medium, and large after removal of superficial tissues. As the larva to specimen ratio strongly influences cleaning time and effectiveness, we first standardized our setup by processing all specimens in identical containers, each with approximately 700 grams of superworms, allowing us to assess the optimal ratio. Our results showed that a high larva to specimen ratio led to damage of fragile bones, while lower ratios resulted in increased cleaning times. Through multiple trials, we suggest that a larva to specimen ratio of 10–15 balances efficient cleaning with minimal risk of bone damage. Applying this ratio to additional bird skulls resulted in thorough cleaning with no observed bone damage. Superworms removed soft tissues within hours to days, depending on specimen size, and were able to clean internal cavities that are typically difficult to reach. Unlike dermestid beetle colonies, which include multiple life stages and pose a higher risk of infestation or egg dispersal, superworms are limited to the larval stage, reducing such risks. Our findings demonstrate that superworms offer a rapid, adaptable, and museum-safe alternative to conventional skeletal cleaning methods, providing an efficient and practical option for scientific and curatorial settings. Superworms are readily available from commercial breeders, and maintaining a colony is straightforward, further supporting their use as a viable alternative for skeletal preparation.
Synergistic Electron‐Proton Transfer Over In <sub>2</sub> O <sub>3</sub> /CuGa <sub>0.5</sub> S Z‐Scheme Heterojunction for Highly Selective CO <sub>2</sub> ‐to‐CH <sub>4</sub> Photoconversion
ABSTRACT Solar‐driven photoconversion of CO 2 and H 2 O into value‐added chemicals such as CH 4 remains a promising yet challenging strategy, hindered by inefficient charge separation and sluggish proton migration kinetics. Herein, an interfacial‐engineered Z‐scheme In 2 O 3 /CuGa 0.5 S heterojunction is designed to synchronize electron transfer with proton delivery. Utilizing Kelvin probe force microscopy, a direct Z‐scheme charge‐transfer pathway with electron accumulation on CuGa 0.5 S is identified. Electronic‐state modulation facilitating CO 2 activation is revealed by quasi‐ in situ XANES and operando XPS, while water dissociation is promoted by In–O sites on In 2 O 3 for continuous reactive H* supply. Preferential H* relocation to CuGa 0.5 S and coupling with *COOH intermediates are demonstrated by in situ DRIFTS and DFT calculations, through which continuous hydrogenation toward CH 4 is driven. Key intermediates are stabilized by synergistic interactions between adjacent components, resulting in significantly enhanced CH 4 selectivity and effectively suppressed competing H 2 evolution. Consequently, a CH 4 evolution rate of 319.2 µmol g −1 h −1 with approximately 100% selectivity is achieved over optimized In 2 O 3 /0.5CuGa 0.5 S in pure water. A mechanistic understanding of electron‐coupled proton transfer in photocatalytic CO 2 reduction is provided, offering an efficient pathway for advancing solar‐driven hydrocarbon production technologies.
Landscape configuration shapes cultural ecosystem service perception and interrelationships in the urban parks of Lhasa
Abstract A central challenge in sustainable urban design is quantifying how landscape spatial configuration influences public perception of cultural ecosystem services (CES), particularly in ecologically sensitive high-altitude regions. This exploratory study explored the relationships between landscape configuration and CES perception in Lhasa’s urban parks, adopting text mining of social media comments( $$n=3319$$ ), landscape pattern analysis, spatial mapping, and questionnaire validation. Six CES dimensions were identified, with significant variations in service provision across Parks. Statistical correlation analysis suggested that specific landscape metrics were meaningfully related to CES perception: the Largest Patch Index (LPI) exhibited a strong positive linear relationship with recreation perception ( $$R^2=0.713$$ ), while the Landscape Shape Index (LSI) showed a quadratic trend with aesthetic perception. Spatial analysis revealed distinct spatial patterns for each perceived service. Field surveys supported these exploratory linkages, with mean agreement scores ranging from 3.87 to 4.52 on a 5-point Likert scale. Correlation analysis among CES dimensions revealed negative associations indicative of perceptual interrelationships: Recreation was negatively associated with heritage and culture ( $$r=-0.75$$ ) and education ( $$r=-0.82$$ ), while heritage and culture and education were positively associated ( $$r=0.90$$ ), reflecting synergy. These findings suggest that landscape configuration shapes multidimensional CES perceptions through distinct pathways, forming spatial-functional interactions among perceived services. This study provides exploratory empirical evidence on the impact of landscape configuration on CES perception in high-altitude cities, reveals the interrelationships within perceived service bundles, and offering insights for Park design and management aimed at balancing diverse service needs.
The evolving landscape of CDK inhibitor use in breast cancer therapy and beyond
Bioinspired Super‐Resolution Electrochemiluminescence Imaging Based on Triboelectric Nanogenerator for Electrode‐Electrolyte Interface Analysis
ABSTRACT Understanding the nature of electrode–electrolyte interface remains a long‐sought goal in electrochemistry, as it could help to address major societal challenges of our time, such as the global energy and environmental crisis. Electrochemiluminescence (ECL) imaging enables the direct visualization of the electrode–electrolyte interface, rather than inferring active sites from morphology or surface chemistry. However, the signal decay and limited stability of ECL signals restrict their practical application. Here, by utilizing the high‐frequency pulsed output of a triboelectric nanogenerator (TENG), we developed a TENG‐based electrochemiluminescence (TECL) imaging inspired by the cardiac diastole‐systole cycle. The TECL signal overcomes the limitations of traditional ECL and employs a zero‐shot deconvolution network (ZS‐DeconvNet) to achieve super‐resolution TECL (ZSSR‐TECL) imaging. Utilizing the ZSSR‐TECL imaging, the catalytic regions at the electrode‐electrolyte interface can be analyzed, with high‐activity, moderate‐activity, and degraded regions clearly distinguished. In addition, three methods have been established for analyzing the catalytic activity of the electrode‐electrolyte interface, including direct observation of TECL images, analysis of signal intensity profiles, and quantification of total signal intensity in fixed regions. The TECL imaging has the potential to enable the research of electrode‐electrolyte interfaces and offers a perspective for enhancing ECL imaging.
Cytosolic heat shock protein 90 is required for photoreceptor outer segment development and vision
Stage distribution and survival outcome in a real-world cervical cancer cohort from an urban-dominant population in Indonesia.
219 Background: Cervical cancer is the second most common female malignancy in Indonesia and remains one of the leading causes of cancer morbidity and mortality in low- and middle-income countries. While urban areas are commonly associated with greater community awareness and access to healthcare than their rural counterparts, real-world data on stage distribution and survival outcome comparisons between these two areas are still limited. This study aimed to describe the clinicopathological distribution and short-term survival outcomes in a real-world cervical cancer cohort in one of the cancer referral centers in Indonesia. Methods: This retrospective observational study included cervical cancer patients treated at a high-volume oncology center in North Sumatra, Indonesia, between January and December 2024. Patient data, such as clinicopathologic characteristics and residential location, were collected from medical records. Survival outcomes were evaluated and compared using Kaplan-Meier estimation and the log-rank test. Results: This study included 162 patients with a median age of 53 years, ranging between 29 and 89 years. The majority of patients had squamous cell carcinoma (83.2%), an ECOG performance status of 0-1 (92.6%), and lived in urban settings (63.6%). The 116 patients with available staging data consisted of 63.8% who exhibited stage III disease and 20.7% with stage IV disease, indicating a significant prevalence of advanced-stage disease. Notably, stage IV patients were dominated by urban residents (62.5%). The percentage of distant metastasis (stage IVB) was comparable between urban and rural patients, at 10.7% and 10.1%, respectively. In terms of treatment completion, only 84% of the 100 patients intended to receive concurrent chemoradiation and brachytherapy completed their treatment. During a median follow-up period of 7.3 months (95% CI 6.4–8.3), 27 patients (16.7%) died and 19 (11.7%) showed disease progression. The one-year overall survival (OS) rate was 66.7% for stage IVA and 57% for stage IVB cervical cancer (log-rank p=0.017). Conclusions: Despite the predominance of urban residents within our real-world cohort, a large proportion of patients presented with late-stage disease. Our findings suggest persistent delays in diagnosis, even in settings where better healthcare accessibility is expected. Health system gaps surrounding public awareness, early detection, and timely diagnosis serve as targets for national health system reform, regardless of location.
The AI–CDK4/6i synergy matrix: A de-novo meta-analysis decoupling aromatase-inhibitor identity from survival outcomes in 1L HR+/HER2− advanced breast cancer.
59 Background: Clinical guidelines currently treat letrozole (LET) and anastrozole (ANA) as interchangeable non-steroidal aromatase inhibitor (NSAI) backbones for CDK4/6 inhibitors (CDK4/6i). However, the marked divergence in overall survival (OS) across pivotal trials—despite nearly identical progression-free survival (PFS)—suggests that specific agent combinations may possess unique biological synergies. This study is the first de-novo meta-analysis to decouple AI identity from CDK4/6i selection, aiming to identify the optimal therapeutic matrix for 1L metastatic disease. Methods: We performed a systematic investigation of pivotal 1L RCTs (PALOMA-2, MONALEESA-2, MONARCH-3) comparing CDK4/6i + AI vs. AI alone. Utilizing fixed-effect pooling of published Hazard Ratios (HR) and 95% Confidence Intervals (CI), we calculated "Survival Efficiency" for Letrozole-fixed backbones versus permissive-NSAI backbones (which include ANA). We specifically modeled the "PFS-to-OS Conversion Ratio" to detect signal-intensity across clinical phenotypes, including high-risk visceral and liver metastases. Results: Pooled analysis confirms a rigid class effect for PFS in letrozole-backbone trials (HR 0.574, 95% CI: 0.49–0.67), mirroring the mixed-NSAI effect in MONARCH-3 (HR 0.535). However, OS outcomes diverged sharply by inhibitor agent rather than AI partner. Ribociclib + LET demonstrated a definitive OS benefit (HR 0.76), while Palbociclib + LET remained neutral (HR 0.96). Abemaciclib + NSAI showed a strong numerical trend (HR 0.804). Our computed pooled OS HR for a letrozole-only backbone was 0.85 (95% CI: 0.736–0.977), a result driven by Ribociclib’s performance. Conclusions: This analysis identifies a critical "Synergy Gap" where 1L PFS gains do not guarantee OS extension. We conclude that AI identity is a passive clinical substrate; the specific CDK4/6i agent is the master regulator of survival. Ribociclib + Letrozole represents the current optimal OS-validated pairing. Clinicians should prioritize specific inhibitor OS data, particularly in high-risk visceral phenotypes, rather than assuming AI-partner interchangeability. These findings suggest a need to refine guidelines to reflect agent-specific survival hierarchies. Pooled OS and PFS estimates by therapeutic matrix. Cohort n OS HR (95% CI) PFS HR (95% CI) Ribociclib + Letrozole 668 0.76 (0.63–0.93) 0.56 (0.45–0.70) Palbociclib + Letrozole 666 0.96 (0.78–1.18) 0.58 (0.46–0.72) Abemaciclib + NSAI 493 0.80 (0.64–1.01) 0.53 (0.42–0.66) Pooled Letrozole-Backbone 1334 0.85 (0.74–0.98) 0.57 (0.49–0.67)
Achieving near-theoretical strength and high elasticity in micrometer scale TiB2 ceramics
Abstract Achieving near-theoretical strength and elastic limits in crystalline solids remains challenging, yet defect sensitivity typically restricts such behavior to nanometer scale specimens. Here we report experimental evidence of near-theoretical strength and large elastic tensile strains in micrometer scale TiB 2 ceramics, produced in situ by eutectic solidification in steel. In situ bending of crystallographically oriented cantilevers, fixed end beams and C-shaped structures, combined with specimen specific finite element analysis, reveals tensile side stresses of tens of gigapascals and elastic strains up to 9%. Direct microscale tension shows that a large gauge volume exceeding 4 μm 3 sustains nearly uniform tensile stress of ~13 GPa without fracture before grip edge failure, providing a conservative lower bound tensile benchmark, while micropillar compression confirms high stress bearing capability. These results establish eutectic solidification as a scalable pathway to suppress strength limiting defects over micrometer scale volumes, extending near-theoretical ceramic strength beyond the nanoscale and enabling robust microarchitected components.
Ultrafast X-ray Pump–Probe Investigation of the Formation Dynamics of SiV Centers in Diamond
Analysis of the current situation of female induced abortion under China’s “three-child policy”
Objective To assess the changes in the characteristics of women undergoing induced abortion and their contraceptive choices before and after the implementation of the “three-child” policy, and to inform strategies for improving reproductive health services in China. Methods We conducted a retrospective comparative study at a tertiary maternity hospital in Fujian, China. Women who underwent induced abortion between 31 May 2019 and 30 May 2020 (pre-policy period) and those who underwent the procedure between 31 May 2023 and 30 May 2024 (post-policy period) were included. We used R software (version 4.3.2) to compare contraceptive use and method choices prior to the abortion, as well as the primary reasons for unintended pregnancy between pre- and post-policy periods. Results In the pre-policy period, non-use of contraception was the leading cause of unintended pregnancy (43.38%), primarily attributed to a perception of low pregnancy risk. Among contraceptive users, condoms were the most common method (26.47%). Post-policy, the rate of contraceptive non-use decreased to 32.19%. However, reliance on low-efficacy methods persisted, with condoms (43.49%) and withdrawal (23.51%) remaining predominant. Conclusion Non-use of contraception and reliance on low-efficacy methods remained the primary contributors to unintended pregnancies among women undergoing induced abortion both before and after the three-child policy. Higher educational attainment was associated with lower contraceptive failure risk. These findings underscore the urgent need to improve access to effective contraceptive methods and strengthen reproductive health education to reduce unintended pregnancies and safeguard women’s reproductive health.
Automatic emotional responses as a transdiagnostic risk and emotional control as a protective factor for adolescent psychopathology: a network analysis
Abstract Adolescence is a critical developmental period marked by heightened risk for emotion-related impulsivity (ERI), challenges in emotion regulation (ER), and mental health difficulties. Dual-process models propose that individuals respond to emotions either automatically or deliberately, with implications for mental health. We applied network analysis using a Gaussian Graphical Model to examine associations between automatic and deliberate emotional responses and symptoms of internalizing, externalizing, and thought problems in a general population of adolescents from south-east Norway ( N = 918; Mage = 13.97, SD = 0.93, range = 12–18; 54.60% female). This cross-sectional study relied on self-report. Automatic ER and ERI nodes exhibited the most connections with symptom nodes across internalizing, externalizing, and thought problems. The strongest association linked limited access to ER strategies with feeling down and depressed. Deliberate ER nodes were also interconnected with symptoms, the strongest association showing more emotional control being linked to less feeling bad or not liking oneself. Several nodes demonstrated many or strong connections with other nodes within and across their groupings (i.e., strength and bridge centrality, respectively). Findings suggest that responding automatically may reflect a transdiagnostic risk factor and emotional control may serve as a protective factor, and these may be particularly useful targets in intervention and prevention efforts. Future studies are needed to determine causality.
Dual tumour–myeloid targeting of glioblastoma with GPNMB CAR-T cells
Abstract Glioblastoma is a lethal brain tumour for which current multimodal treatment rarely prevents recurrence 1 . Therapeutic failure is driven by extensive intratumoural cellular heterogeneity 2 with a microenvironment dominated by tumour-associated macrophages that sustain tumour growth and immunosuppression 3 . Although chimeric antigen receptor (CAR)-T cell therapies are being developed for glioblastoma, sustained response has been undermined by non-uniform antigen expression, antigen loss and microenvironmental barriers that are not directly engaged by tumour-targeting designs 4 . These limitations motivate new strategies that address the disease as a coupled tumour–immune system rather than a single malignant compartment. Here we use a multi-omic target discovery platform to identify GPNMB as a dual-compartment antigen in glioblastoma. Anti-GPNMB CAR-T cells showed potent anti-tumour activity, with long-term disease control in orthotopic patient-derived xenografts and syngeneic glioma models through concomitant depletion of GPNMB + tumour and immunosuppressive myeloid populations. By collapsing tumour control and microenvironmental reprogramming, these findings provide a new strategy for antigen selection and targeting in heterogenous, myeloid-rich solid cancers.
Wearable Flexible Sensors for Cardiovascular Disease Monitoring
ABSTRACT Wearable flexible sensors have emerged as a cornerstone of next‐generation bioelectronics, enabling skin‐conformal, continuous, and high‐fidelity monitoring of cardiovascular diseases (CVDs). This review elucidates the structure‐function relationships that govern sensing performance, highlighting how material innovation, structure engineering, and device architectures synergistically balance sensitivity, mechanical robustness, and biocompatibility. Key cardiovascular physiological signals, including electrical, mechanical, hemodynamic, and biochemical modalities, are systematically summarized and correlated with representative sensing mechanisms such as piezoresistive, capacitive, triboelectric, electrochemical, and optical transduction. The integration of machine learning (ML) and data‐driven modeling is further discussed, highlighting its potential to enable personalized diagnostics, multimodal fusion, and adaptive prediction of cardiovascular risks. Despite substantial progress, critical challenges remain in long‐term operational stability, scalable manufacturing, cross‐population generalizability, and clinical validation. To address these limitations, a unified design paradigm integrating materials engineering, multimodal sensing strategies, and algorithmic intelligence is proposed. This review aims to guide the development of next‐generation wearable platforms that are not only mechanically compliant and functionally robust but also algorithmically interpretable and clinically translatable, laying the groundwork for intelligent, reliable, and precision‐oriented CVD monitoring systems.
A prokaryotic-like cardiolipin synthase is essential to maintain mitochondrial function, lipid homeostasis, and survival of the blood stage malaria parasite Plasmodium falciparum
Real world outcomes with neoadjuvant chemo-immuno-metronomic therapy in advanced head and neck squamous cell carcinoma (NeoLOCUS-Retro).
236 Background: Patients with stage IVA–IVB head and neck squamous cell carcinoma (HNSCC) ineligible for upfront surgery or definitive chemoradiation have poor 5-year survival of 30% to 40%. The prospective NeoLOCUS study showed high response and conversion rates with a regimen combining chemotherapy, low-dose immunotherapy (LD-IO), and oral metronomic therapy (OMT). Here, we report real-world outcomes with the same. Methods: Single-center retrospective study of stage IVA–IVB HNSCC treated with neoadjuvant platinum–taxane chemotherapy plus LD-IO and OMT (erlotinib, methotrexate, celecoxib). Primary endpoint was conversion to radical-intent local therapy. Secondary endpoints were radiologic response, event-free survival (EFS), overall survival (OS), and grade ≥3 adverse events. Results: Forty-three patients were included (Characteristics in Table). After a median 4 (Range 1- 6) neoadjuvant cycles, objective radiologic response was 62.8% (complete response 2.3%; partial response 60.5%, progressive disease 4.7%). Volumetric tumor reduction ≥65% occurred in 54.8%. Conversion to radical-intent therapy was achieved in 69.8%. At median follow-up of 16 months, median EFS was 10.5 months and OS 16.1 months. Thirteen deaths (30.2%) occurred. Median OS was 18.1 months in those receiving radical therapy versus 6.5 months in those who did not. Grade ≥3 adverse events occurred in 14.7%, most commonly anemia and thrombocytopenia (7% each). One grade 4 diarrhea with acute kidney injury required discontinuation. At disease progression, 20.9% received palliative systemic therapy, 2.3% palliative radiotherapy alone and others supportive care. Conclusions: NeoLOCUS regimen demonstrated encouraging response rates and enabled conversion to radical-intent local therapy in about two-thirds of patients with advanced stage IVA–IVB HNSCC, translating into improved survival in this challenging clinical setting. Baseline and treatment characteristics. Characteristic Value Median age, yr (IQR) 54 (45.5–60) Male sex, n (%) 33 (76.7) Stage IVA / IVB, n (%) 25 (58.1) / 18 (41.9) Oral vs non-oral primary, n (%) 28 (65.1) / 15 (34.9) Harmonised ICI dose mg/kg/2wks (IQR) 0.4 (0.2–0.4) Radical-intent local therapy, n (%) 30 (69.8) CTRT + Radical RT, n (%) 24 (55.8) Surgery, n (%) 6 (14.0) Palliative RT / No local therapy 3 (7.0) / 10 (23.3)
Assessing the clinical and biological associations between multimodal artificial intelligence (MMAI) and 22-gene genomic classifier (GC) in localized prostate cancer (PCa).
196 Background: MMAI and GC are clinically validated prognostic tools for localized PCa, and are incorporated into NCCN guidelines as treatment-decision aids. However, their clinical and biological relationships remain unclear, especially in Asian populations where comparative study is limited. We investigated the association between these genomic and AI pathology biomarkers and their respective accuracies for the prognostication of metastasis risk in patients with NCCN intermediate- (Int) to high-risk PCa. Methods: Patients with newly-diagnosed localized PCa from a single institution in Singapore were enrolled into a prospective protocol (NCT04340024), and underwent image-guided radiotherapy with or without hormonal therapy. All subjects had paired GC and MMAI scores. However, colored marker annotations were present on the H&E slides, which could negatively impact MMAI performance. Associations between GC and MMAI with metastasis-free survival (MFS) were evaluated using Cox proportional hazards models and and Area under the Receiver Operating Characteristic Curve (AUC). Pearson correlation was used to assess associations of cancer hallmark pathways. Results: Of 144 included men (142 NCCN Int/High; median follow-up 78.5 mo), GC and MMAI scores were correlated ( R = 0.55) but differed in risk stratification; GC classified 50/142 (35%) NCCN Int/High cases as GC low-, 31/142 (22%) as int-, and 61/142 (43%) as high-risk, whereas MMAI assigned 11/142 (8%), 53/142 (37%), and 78/142 (55%), respectively. Both tests were prognostic for MFS (MMAI: hazard ratio [ HR ] of 2.6 [95% CI:1.0–6.4], P =0.035; GC: HR of 3.4 [95% CI:1.5–7.6], P =0.002), although the difference in HR should be interpreted with caution, as it is attributable to a limited sample size with longer follow-up. Adding them to NCCN improved 5y MFS AUCs from 0.71 to 0.80 for both NCCN+MMAI and NCCN+GC. Biologically, MMAI-high and GC-high groups showed comparable proportions of aggressive subtypes (PAM50 Basal 51% vs 58%, P =0.419; Luminal B 46% vs 39%, P =0.377; PTEN loss 22% vs 25%; P =0.698; AR lower activity 49% vs 44%, P =0.512), though pathway enrichment diverged; GC high versus GC low tumors showed positive enrichment in E2F, oxidative phosphorylation, and mTORC1 pathways; negative enrichment in P53 and estrogen pathways, whereas MMAI high was uniquely negatively enriched for fatty acid metabolism. Conclusions: Both MMAI and GC are biomarkers that enhance NCCN risk stratification, though they likely capture distinct biological signals. Incorporating these tests with NCCN guidelines may better inform treatment escalation decisions—such as the addition of hormonal therapy for NCCN-Int or anti-androgens for NCCN-High risk disease—in localized PCa.
Molecular basis for depsipeptide HDAC inhibitor combinatorial biosynthesis
Abstract Polyketides and nonribosomal peptides are important natural product classes with wide-ranging medical and agricultural applications. The analogous enzymatic logic employed by bacterial modular polyketide synthases (PKSs) and nonribosomal peptide synthetases (NRPSs) enables the assembly of hybrid products. One important group of polyketide-nonribosomal peptide hybrids is exemplified by the HDAC-targeting drug romidepsin. This group is assembled by combinatorial biosynthesis involving fusion of a conserved Zn 2+ -binding pharmacophore to a variable peptide-based cap. Here, we use gene proximity searching to identify the FR-901375 biosynthetic gene cluster in Pseudomonas chlororaphis subsp. piscium DSM 21509. Comparison of the PKS-NRPS encoded by this gene cluster with those assembling related depsipeptide HDAC inhibitors suggests an unusual subunit docking modality enables interaction between the conserved pharmacophore and variable cap biosynthetic machineries. This hypothesis is validated using crosstalk assays, mutagenesis, AlphaFold predictions, and carbene footprinting, providing insight into the evolution of mechanisms for hybrid polyketide-nonribosomal peptide combinatorial biosynthesis.