Ca(H2PO4)2 and MgSO4 activated nitrogen-related bacteria and genes in thermophilic stage of compost.

This study was conducted to investigate the effects of Ca(H2PO4)2 and MgSO4 on the bacterial community and nitrogen metabolism genes in the aerobic composting of pig manure. The experimental treatments were set up as control (C), 1% Ca(H2PO4)2 + 2% MgSO4 (CaPM1), and 1.5% Ca(H2PO4)2 + 3% MgSO4 (CaPM2), which were used at the end of composting for potting trials. The results showed that Ca(H2PO4)2 and MgSO4 played an excellent role in retaining nitrogen and increasing the alkali-hydrolyzed nitrogen (AN), available phosphorus (AP), and available potassium (AK) contents of the composts. Adding Ca(H2PO4)2 and MgSO4 changed the microbial community structure of the compost. The microorganisms associated with nitrogen retention were activated. The complexity of the microbial network was enhanced. Genetic prediction analysis showed that the addition of Ca(H2PO4)2 and MgSO4 reduced the accumulation of nitroso-nitrogen and the process of denitrification. At the same time, despite the reduction of genes related to nitrogen fixation, the conversion of ammonia to nitrogenous organic compounds was promoted and the stability of nitrogen was increased. Mantel test analysis showed that Ca(H2PO4)2 and MgSO4 can affect nitrogen transformation-related bacteria and thus indirectly affect nitrogen metabolism genes by influencing the temperature, pH, and organic matter (OM) of the compost and also directly affected nitrogen metabolism genes through PO43- and Mg2+. The pot experiment showed that composting with 1.5% Ca(H2PO4)2 + 3% MgSO4 produced the compost product that improved the growth yield and nutrient content of cilantro and increased the fertility of the soil. In conclusion, Ca(H2PO4)2 and MgSO4 reduces the loss of nitrogen from compost, activates nitrogen-related bacteria and genes in the thermophilic phase of composting, and improves the fertilizer efficiency of compost products. KEY POINTS: • Ca(H2PO4)2 and MgSO4 reduced the nitrogen loss and improved the compost effect • Activated nitrogen-related bacteria and altered nitrogen metabolism genes • Improved the yield and quality of cilantro and fertility of soil.

Effect of nitrogen retention composite additives Ca(H2PO4)2 and MgSO4 on the degradation of lignocellulose, compost maturation, and fungal communities in compost.

This study investigated the effects of the nitrogen retention composite additives Ca(H2PO4)2 and MgSO4 on lignocellulose degradation, maturation, and fungal communities in composts. The study included control (C, without Ca(H2PO4)2 and MgSO4), 1% Ca(H2PO4)2 + 2% MgSO4 (CaPM1), 1.5% Ca(H2PO4)2 + 3% MgSO4 (CaPM2). The results showed that Ca(H2PO4)2 and MgSO4 enhanced the degradation of total organic carbon (TOC) and promoted the degradation of lignocellulose in compost, with CaPM2 showing the highest TOC and lignocellulose degradation. Changes in the three-dimensional excitation-emission matrix fluorescence spectroscopy (3D-EEM) of dissolved organic matter (DOM) components in compost indicated that the treatment group with the addition of Ca(H2PO4)2 and MgSO4 promoted the production of humic acids (HAs) and increased the degree of compost decomposition, with CaPM2 demonstrating the highest degree of decomposition. The addition of Ca(H2PO4)2 and MgSO4 modified the composition of the fungal community. Ca(H2PO4)2 and MgSO4 increased the relative abundance of Ascomycota, decreased unclassified_Fungi, and Glomeromycota, and activated the fungal genera Thermomyces and Aspergillus, which can degrade lignin and cellulose during the thermophilic stage of composting. Ca(H2PO4)2 and MgSO4 also increased the abundance of Saprotroph, particularly undefined Saprotroph. In conclusion, the addition of Ca(H2PO4)2 and MgSO4 in composting activated fungal communities involved in lignocellulose degradation, promoted the degradation of lignocellulose, and enhanced the maturation degree of compost.

Effective Motion Self-Learning Genre Using 360° Virtual Reality Content on Mobile Device: A Study Based on Taichi Training Platform.

Online fitness training, with its affordability and flexibility, offers a convenient way for individuals to engage in regular workouts that promote physical and mental health. Yet, learning fitness motions in this way presents various challenges and may not always be as effective as in-person training. To address the practical demands of motion learning, we conducted a systematic survey and accordingly proposed a four-stage self-learning genre that integrates immersive virtual reality (VR) environments with motion skill learning theories, strategies, and expert experience. Herein, we merged progressive structures and multi-level visual cues to enhance instruction, and proposed a fine-grained motion analysis method to provide adaptive correction feedback during training. Utilizing a Taichi training platform with the genre embedded, we systematically validated the effectiveness of the genre, and examined the potential impact of VR content presentation form on motion learning among different age groups, as well as their preferences and focus on VR fitness training genre design. Results from the quantitative analysis, qualitative evaluation, and case study showed that the 360° video-based VR content brought more positive motion learning performance and user experiences than the fully-simulated VR used in many previous studies. The proposed genre demonstrated outstandingly performance, experience, and usability, with each stage and design playing an effective role. Moreover, we offer several design considerations for VR fitness systems targeting diverse age groups, providing beneficial insights for VR development in the sports and health-related fields.

Anti-Inflammatory Terpenoids from the Rhizomes of Shell Ginger.

Shell ginger (Alpinia zerumbet) is a perennial ornamental plant of ginger native to East Asia, which can be used as a flavoring agent in food or beverage, as well as a traditional Chinese medicine. In this study, a total of 37 terpenoids, including 7 new compounds, zerumin D1 to zerumin D7 (2, 3, 28-30, 36, and 37), and 5 new naturally occurring compounds, zerumin D10 to zerumin D14 (9, 12, 15, 20, and 24), were isolated and identified from the rhizomes of shell ginger. Compound 3 was an unprecedented variant labdane diterpenoid featuring a unique 6/7/6/3 tetracyclic cyclic ether system in its side chain. The anti-inflammatory activities of the isolated terpenoids were assessed in RAW 264.7 macrophages stimulated by lipopolysaccharide (LPS). Compound 4 significantly inhibited the production of nitric oxide with an IC50 value of 5.4 µM. Further investigation revealed that compounds 2 and 3 may inhibit the nuclear translocation of NF-κB, thus suppressing the expression of IL-6, IL-1ß, iNOS, and COX-2 to exert the anti-inflammatory effects.

Anti-proliferative and anti-inflammatory eudesmanolides from the flowers of Sphagneticola trilobata (L.) Pruski.

Twenty-six eudesmanolides including six undescribed compounds were isolated from the flowers of Sphagneticola trilobata (L.) Pruski. Their structures were elucidated based on the interpretation of spectroscopic techniques, NMR calculation, and DP4+ analysis. The stereochemistry of (1S,4S,5R,6S,7R,8S,9R,10S,11S)-1,4,8- trihydroxy-6-isobutyryloxy-11-methyleudesman-9,12-olide (1) was demonstrated by single crystal X-ray diffraction. All eudesmanolides were evaluated for their anti-proliferative activities against four human tumor cell lines (HepG2, HeLa, SGC-7901, and MCF-7). 1α,4ß-Dihydroxy-6α-methacryloxy-8ß-isobutyryloxyeudesman-9,12-olide (3) and wedelolide B (8) showed pronounced cytotoxic effects against AGS cell line with IC50 values of 1.31 and 0.89 µM, respectively. Their anti-proliferative activities against AGS cells were exerted through a dose-dependent apoptosis pathway, as verified by cell and nucleus morphological assessment, clone formation assay, and Western blot analysis. Furthermore, 1α,4ß,8ß-trihydroxy-6ß-methacryloxyeudesman-9,12-olide (2) and 1α,4ß,9ß-trihydroxy-6α-isobutyryloxy- 11α-13-methacryloxyprostatolide (7) performed significant inhibitory effects on lipopolysaccharide-stimulated nitric oxide production in RAW 264.7 macrophages with IC50 values of 11.82 and 11.05 µM, respectively. Moreover, compounds 2 and 7 could block the nuclear translocation of NF-κB and reduce the expression of iNOS, COX-2, IL-1ß, and IL-6 to exert anti-inflammatory effects. This study provides evidence for the utilization of the eudesmanolides from S. trilobata as lead compounds for further research due to their cytotoxic potential.

Biomimetic cell membrane-coated nanocarriers for targeted siRNA delivery in cancer therapy.

Small interfering RNA (siRNA) therapeutics for cancer are a focus of increasing research interest. However, the major obstacle to their clinical application is the targeted delivery of siRNA to cancer cells at desirable levels. Cell membrane-coated nanocarriers have the advantage of combining the properties of both cell membranes and nanoparticles (NPs). In this review, we highlight the most common RNAi therapeutics and the extracellular and intracellular barriers to siRNA delivery. Moreover, we discuss clinical applications of different cell membrane-coated nanocarriers for targeted siRNA delivery, including cancer cell membranes (CCMs), platelet membranes, erythrocyte membranes, stem cell membranes, exosome membranes, and hybrid membranes. Taken together, biomimetic cell membrane-coated nanotechnology is a promising strategy for targeted siRNA delivery for cancer treatment.

Synergistic Combination of the Total Steroidal Saponins from the Berries of Black Nightshade and Adriamycin to Overcome Leukemia Multidrug Resistance.

The berries of black nightshade (Solanum nigrum L.) are consumed as a favorite fruit in some regions and have been reported to possess a range of biological activities. Previous studies have found that the steroidal saponins from the berries of S. nigrum (SN) showed potential antileukemic activity, although the underlying mechanism remains to be revealed. This study investigated the effects and mechanisms of SN in combination with adriamycin to reverse leukemia multidrug resistance in vivo and in vitro. The results indicated that the combination of SN and adriamycin displayed enhanced suppression ability both in vitro and in vivo by the modulation of drug efflux proteins. Further study revealed that SN and adriamycin co-treatment induced cell apoptosis in K562/ADR cells through caspase pathways and autophagy through the PI3K/Akt/mTOR and MAPK signaling pathway. This study provides a new prospect of the berries of black nightshade in multidrug resistance therapy of cancer.

Macrophage membrane-coated nanovesicles for dual-targeted drug delivery to inhibit tumor and induce macrophage polarization.

Background: Immunosuppressive M2 macrophages in the tumor microenvironment (TME) can mediate the therapeutic resistance of tumors, and seriously affect the clinical efficacy and prognosis of tumor patients. This study aims to develop a novel drug delivery system for dual-targeting tumor and macrophages to inhibit tumor and induce macrophage polarization. Methods: The anti-tumor effects of methyltransferase like 14 (METTL14) were investigated both in vitro and in vivo. The underlying mechanisms of METTL14 regulating macrophages were also explored in this study. We further constructed the cyclic (Arg-Gly-Asp) (cRGD) peptide modified macrophage membrane-coated nanovesicles to co-deliver METTL14 and the TLR4 agonist. Results: We found that METTL14 significantly inhibits the growth of tumor in vitro. METTL14 might downregulate TICAM2 and inhibit the Toll-like receptor 4 (TLR4) pathway of macrophages, meanwhile, the combination of METTL14 and the TLR4 agonist could induce M1 polarization of macrophages. Macrophage membrane-coated nanovesicles are characterized by easy modification, drug loading, and dual-targeting tumor and macrophages, and cRGD modification can further enhance its targeting ability. It showed that the nanovesicles could improve the in vivo stability of METTL14, and dual-target tumor and macrophages to inhibit tumor and induce M1 polarization of macrophages. Conclusions: This study anticipates achieving the dual purposes of tumor inhibition and macrophage polarization, and providing a new therapeutic strategy for tumors.

Dapagliflozin-Loaded Exosome Mimetics Facilitate Diabetic Wound Healing by HIF-1α-Mediated Enhancement of Angiogenesis.

Angiogenesis plays a critical role in diabetic wound healing. However, no effective strategies have been developed to target endothelial cells (ECs) to facilitate diabetic wound healing. Dapagliflozin (DA) as a sodium-glucose linked transporter 2 (SGLT2) inhibitor, may promote neovascularization in diabetic mice via HIF-1α-mediated enhancement of angiogenesis. Here, the bioinspired nanovesicles (NVs) prepared from induced pluripotent stem cells-derived ECs through an extrusion approach are reported, which can function as exosome mimetics to achieve targeted deliver of DA. Abundant membrane C-X-C motif chemokine receptor 4 conferred the EC-targeting ability of these NVs and the endothelial homology facilitated the accumulation in ECs. Furthermore, these DA-loaded induced pluripotent stem cells (iPSC)-EC NVs can facilitate angiogenesis and diabetic wound healing by HIF-1α/VEGFA pathway. Taken together, this study indicated that targeting ECs and regulating angiogenesis may be a promising strategy for the treatment of diabetic wound healing.

Systematic pan-cancer analysis of the nicotinamide n-methyltransferase in human cancer.

In several tumors, Nicotinamide N-Methyltransferase (NNMT) was identified as a bridge between methylation metabolism and tumorigenesis and was associated with a poor prognosis. This research aims is to study the prognostic value of NNMT in cancer, its relationship with DNA methylation, and the immune microenvironment. On the basis of the Cancer Genome Atlas and the Cancer Cell Line Encyclopedia, Genotype Tissue-Expression, cBioPortal, Cellminer, Gene Expression Profiling Interactive Analysis, Human Protein Atlas and Clinical Proteomic Tumor Analysis Consortium, we used a series of bioinformatics strategies to investigate the potential carcinogenicity of NNMT, including the relationship between NNMT expression and prognosis, tumor mutational burden, microsatellite instability, and sensitivity analysis of anticancer drugs. The GeneMANIA, STRING, and BioGRID databases were examined for protein-protein interactions, and Gene Ontology and the Kyoto Encyclopedia of Genes were used to infer the signal pathway. The results indicated that NNMT was significantly expressed in several tumor tissues compared to the matching non-tumor tissues. Increased NNMT expression was linked to reduced OS, DSS, and DFI. In addition, there was a link between NNMT expression and TMB and MSI in 18 cancer types, and between NNMT expression and DNA methylation in 23 cancer types. Further study of NNMT gene alteration data revealed that deletion was the most prevalent form of NNMT mutation, and that there was a significant negative association between NNMT expression and mismatch repair genes. In addition, there was a strong positive connection between NNMT and immune infiltration in 28 types of tumors, and the immune cells that infiltrated the tumors displayed a characteristic NNMT pattern. According to the enrichment study, cell migration, cell motility, and cell adhesion were highly enriched in biological processes, and NNMT may be associated with the PI3K-Akt signaling pathway. By downregulating gene methylation or impacting the immunological microenvironment widely, NNMT may drive carcinogenesis and cause a poor prognosis. Our research showed that NNMT could be used as a biomarker of tumor immune infiltration and poor prognosis, thus providing a unique strategy for cancer therapy.

Biomimetic cell membrane-coated glucose/oxygen-exhausting nanoreactor for remodeling tumor microenvironment in targeted hypoxic tumor therapy.

Hypoxia is a common feature within many types of solid tumors, which is closely associated with limited efficacy for tumor therapies. Moreover, the inability to reach hypoxic tumor cells that are distant from blood vessels results in tumor-targeting and penetrating drug delivery systems in urgent need. Here, glucose oxidase (GOX) and hypoxia-activated prodrug tirapazamine (TPZ) are loaded into photothermal conversion agent polydopamine (PDA) as the glucose/oxygen-exhausting nanoreactor named PGT. We further construct a tumor cell membrane-coated nanovesicle for the targeted delivery of PGT. This biomimetic nanovesicle exhibits significantly improved tumor-targeting and tumor-penetrating abilities. After internalization by the tumor cells, the loaded drug is quickly released in response to near-infrared (NIR) laser. The PGT nanoreactor can exhaust glucose and oxygen, and further enhance hypoxia within tumor, which efficiently inhibits hypoxic tumor by combining starvation therapy and hypoxia-activated chemotherapy. Mechanically, it is revealed that the nanoreactor significantly increases hypoxia level and downregulates the expression of hypoxia-inhibitory factor-1α (HIF-1α), thereby promoting T cell activation and macrophage polarization to remodel tumor immunosuppressive microenvironment. Therefore, this tumor microenvironment-regulable nanoreactor with sustainable and cascade targeted starvation-chemotherapy provides a novel insight into the treatment of hypoxic tumor.

Engineered biomimetic nanoreactor for synergistic photodynamic-chemotherapy against hypoxic tumor.

Photodynamic therapy (PDT) can produce a large amount of reactive oxygen species (ROS) in the radiation field to kill tumor cells. However, the sustainable anti-tumor efficacy of PDT is limited due to the hypoxic microenvironment of tumor. In this study, classic PDT agent indocyanine green (ICG) and hypoxia-activated chemotherapeutic drug tirapazamine (TPZ) were loaded on mesoporous polydopamine (PDA) to construct PDA@ICG-TPZ nanoparticles (PIT). Then, PIT was camouflaged with cyclic arginine-glycine-aspartate (cRGD) modified tumor cell membranes to obtain the engineered membrane-coated nanoreactor (cRGD-mPIT). The nanoreactor cRGD-mPIT could achieve the dual-targeting ability via tumor cell membrane mediated homologous targeting and cRGD mediated active targeting. With the enhanced tumor-targeting and penetrating delivery system, PIT could efficiently accumulate in hypoxic tumor cells and the loaded drugs were quickly released in response to near-infrared (NIR) laser. The nanoreactor might produce cytotoxic ROS under NIR and further enhance hypoxia within tumor to activate TPZ, which efficiently inhibited hypoxic tumor by synergistic photodynamic-chemotherapy. Mechanically, hypoxia-inhibitory factor-1α (HIF-1α) was down-regulated by the synergistic therapy. Accordingly, the cRGD-mPIT nanoreactor with sustainable and cascade anti-tumor effects and satisfied biosafety might be a promising strategy in hypoxic tumor therapy.

Single-cell transcriptomics reveals the regulative roles of cancer associated fibroblasts in tumor immune microenvironment of recurrent osteosarcoma.

Rationale: Osteosarcoma (OS) is the most common primary bone tumor with a poor prognosis, but the detailed mechanism is still unclear. A comprehensive investigation of tumor microenvironment (TME) of OS might help find effective anti-tumor strategies. Single-cell transcriptomics is a powerful new tool to explore TME. Therefore, this study is designed to investigate the TME and gene expression pattern of primary and recurrent OS at the single-cell level. Methods: The single-cell RNA sequencing and bioinformatic analysis were conducted to investigate the cellular constitution of primary, recurrent, and lung metastatic OS lesions according to the datasets of GSE152048 and GSE162454. TIMER database was used to investigate the role of LOX in the prognosis of sarcoma. The functions of related cells and markers were further confirmed by in vitro and in vivo experiments. Results: Cancer associated fibroblasts (CAFs) were found with a higher infiltrating level in recurrent OS, and were enriched in the epithelial-mesenchymal transition (EMT) pathway. CAFs showed remarkably increased expression of LOX, which might lead to EMT and poor prognosis of OS. Mechanically, LOX regulated the function of CAFs and macrophage polarization to remodel the tumor immune microenvironment. Moreover, LOX inhibitor could inhibit migration and promote apoptosis of OS both in vitro and in vivo. Conclusions: This study revealed the heterogeneity of recurrent OS and highlighted an innovative mechanism that CAFs regulate EMT of OS via LOX. Targeting LOX of CAFs showed promising efficacy in remodeling TME and treating recurrent OS.

Recent advances in crosstalk between N6-methyladenosine (m6A) modification and circular RNAs in cancer.

N6-methyladenosine (m6A), as the most common RNA modification, plays a vital role in the development of cancers. Circular RNAs (circRNAs) are a class of single-stranded covalently closed RNA molecules. Recently, m6A modification has been identified as performing biological functions for regulating circRNAs. Increasing evidence also shows that circRNAs are involved in cancer progression by targeting m6A regulators. In this review, we describe the functional crosstalk between m6A and circRNAs, and illustrate their roles in cancer development. m6A methylation mediates the biogenesis, stability, and cytoplasmic export of circRNAs in different cancer types. Moreover, circRNAs regulate the expression of m6A regulators, participate in the degradation of m6A regulators, and regulate the m6A modification of target mRNAs. Finally, we discuss the potential applications and future research directions of m6A modification and circRNAs in cancer. Further understanding of the biological roles of m6A and circRNAs will provide new insight into the diagnosis and treatment of cancer patients.

Engineered exosome as targeted lncRNA MEG3 delivery vehicles for osteosarcoma therapy.

Exosomes as nanosized membrane vesicles, could targeted deliver therapeutic agents by modification with target ligands. Exosome-derived non-coding RNAs play a vital role in the development of tumors. Previous evidences reveal that long non-coding RNA maternally expressed gene 3 (lncRNA MEG3) has anti-tumor properties. Whereas, the inhibitory effects of exosome-derived lncRNA MEG3 in osteosarcoma (OS) remain largely unknown. In this study, we utilize the engineering technology to combine exosome and lncRNA for tumor-targeting therapy of OS. We elucidated the anti-OS effects of lncRNA MEG3, and then prepared the c(RGDyK)-modified and MEG3-loaded exosomes (cRGD-Exo-MEG3). The engineered exosomes cRGD-Exo-MEG3 could deliver more efficiently to OS cells both in vitro and in vivo. In this way, cRGD-Exo-MEG3 facilitate the anti-OS effects of MEG3 significantly, with the help of enhanced tumor-targeting therapy. This study elucidates that engineered exosomes as targeted lncRNA MEG3 delivery vehicles have potentially therapeutic effects for OS.

Engineered microorganism-based delivery systems for targeted cancer therapy: a narrative review.

Microorganisms with innate and artificial advantages have been regarded as intelligent drug delivery systems for cancer therapy with the help of engineering technology. Although numerous studies have confirmed the promising prospects of microorganisms in cancer, several problems such as immunogenicity and toxicity should be addressed before further clinical applications. This review aims to investigate the development of engineered microorganism-based delivery systems for targeted cancer therapy. The main types of microorganisms such as bacteria, viruses, fungi, microalgae, and their components and characteristics are introduced in detail. Moreover, the engineering strategies and biomaterials design of microorganisms are further discussed. Most importantly, we discuss the innovative attempts and therapeutic effects of engineered microorganisms in cancer. Taken together, engineered microorganism-based delivery systems hold tremendous promise for biomedical applications in targeted cancer therapy.

Exosomes as Efficient Nanocarriers in Osteosarcoma: Biological Functions and Potential Clinical Applications.

Osteosarcoma is the most common bone tumor affecting both adolescents and children. Although localized osteosarcoma has an overall survival of >70% in the clinic, metastatic, refractory, and recurrent osteosarcoma have poorer survival rates. Exosomes are extracellular vesicles released by cells and originally thought to be a way for cells to discard unwanted products. Currently, exosomes have been reported to be involved in intercellular cross-talk and induce changes in cellular behavior by transferring cargoes (proteins, DNA, RNA, and lipids) between cells. Exosomes regulate osteosarcoma progression, and processes such as tumorigenesis, proliferation, metastasis, angiogenesis, immune evasion, and drug resistance. Increasing evidences shows that exosomes have significant potential in promoting osteosarcoma progression and development. In this review, we describe the current research status of exosomes in osteosarcoma, focusing on the biological functions of osteosarcoma exosomes as well as their application in osteosarcoma as diagnostic biomarkers and therapeutic targets.

Sex differences in noradrenergic modulation of attention and impulsivity in rats.

RATIONALE: Noradrenaline (NE) is closely related to attentive performance and impulsive control. However, the potential sex differences regarding attention and impulsivity under the noradrenergic modulation have been largely neglected. Therefore, our study aimed to investigate whether male and female rats exhibit differential responses to NE-related drugs during the five-choice serial reaction time task (5CSRT). METHODS: Male and female rats were trained in 5CSRT and administered with different NE drugs after obtaining stable baseline performance: atipamezole, a highly selective α2 receptor antagonist; prazosin, an α1 receptor antagonist; and atomoxetine, a selective NE reuptake inhibitor. Later, prazosin was selected to co-administration with atomoxetine. RESULTS: Male and female rats exhibited equal learning speed, and no significant baseline differences were found as measured by the 5CSRT. Atomoxetine decreased premature responses in both sexes, but the extent of this reduction was different, with the reduction greater in males. Besides, atomoxetine (1.8 mg/kg) increased the error of omissions in females. The high dose of prazosin (0.5 mg/kg) decreased the accuracy only in male rats, but this was ameliorated by the co-administration with atomoxetine. CONCLUSIONS: Atomoxetine showed significant improvement in impulsivity, but atomoxetine had less beneficial effects on impulsive control in females than in males, and it even impaired attentional performance in female rats. The α1 receptors were mainly responsible for NE drug-related sex differences in attention rather than impulsivity. The results obtained in this study indicate that the sex differences exist in both attention and impulsivity by the modulation of noradrenaline and raise the concern to improve sex-specific treatments.

Adverse Drug Reaction Predictions Using Stacking Deep Heterogeneous Information Network Embedding Approach.

Inferring potential adverse drug reactions is an important and challenging task for the drug discovery and healthcare industry. Many previous studies in computational pharmacology have proposed utilizing multi-source drug information to predict drug side effects have and achieved initial success. However, most of the prediction methods mainly rely on direct similarities inferred from drug information and cannot fully utilize the drug information about the impact of protein⁻protein interactions (PPI) on potential drug targets. Moreover, most of the methods are designed for specific tasks. In this work, we propose a novel heterogeneous network embedding approach for learning drug representations called SDHINE, which integrates PPI information into drug embeddings and is generic for different adverse drug reaction (ADR) prediction tasks. To integrate heterogeneous drug information and learn drug representations, we first design different meta-path-based proximities to calculate drug similarities, especially target propagation meta-path-based proximity based on PPI network, and then construct a semi-supervised stacking deep neural network model that is jointly optimized by the defined meta-path proximities. Extensive experiments with three state-of-the-art network embedding methods on three ADR prediction tasks demonstrate the effectiveness of the SDHINE model. Furthermore, we compare the drug representations in terms of drug differentiation by mapping the representations into 2D space; the results show that the performance of our approach is superior to that of the comparison methods.