iGWAS: Image-based genome-wide association of self-supervised deep phenotyping of retina fundus images.

Existing imaging genetics studies have been mostly limited in scope by using imaging-derived phenotypes defined by human experts. Here, leveraging new breakthroughs in self-supervised deep representation learning, we propose a new approach, image-based genome-wide association study (iGWAS), for identifying genetic factors associated with phenotypes discovered from medical images using contrastive learning. Using retinal fundus photos, our model extracts a 128-dimensional vector representing features of the retina as phenotypes. After training the model on 40,000 images from the EyePACS dataset, we generated phenotypes from 130,329 images of 65,629 British White participants in the UK Biobank. We conducted GWAS on these phenotypes and identified 14 loci with genome-wide significance (p<5×10-8 and intersection of hits from left and right eyes). We also did GWAS on the retina color, the average color of the center region of the retinal fundus photos. The GWAS of retina colors identified 34 loci, 7 are overlapping with GWAS of raw image phenotype. Our results establish the feasibility of this new framework of genomic study based on self-supervised phenotyping of medical images.

Advanced gene nanocarriers/scaffolds in nonviral-mediated delivery system for tissue regeneration and repair.

Tissue regeneration technology has been rapidly developed and widely applied in tissue engineering and repair. Compared with traditional approaches like surgical treatment, the rising gene therapy is able to have a durable effect on tissue regeneration, such as impaired bone regeneration, articular cartilage repair and cancer-resected tissue repair. Gene therapy can also facilitate the production of in situ therapeutic factors, thus minimizing the diffusion or loss of gene complexes and enabling spatiotemporally controlled release of gene products for tissue regeneration. Among different gene delivery vectors and supportive gene-activated matrices, advanced gene/drug nanocarriers attract exceptional attraction due to their tunable physiochemical properties, as well as excellent adaptive performance in gene therapy for tissue regeneration, such as bone, cartilage, blood vessel, nerve and cancer-resected tissue repair. This paper reviews the recent advances on nonviral-mediated gene delivery systems with an emphasis on the important role of advanced nanocarriers in gene therapy and tissue regeneration.

An improved cell line-derived xenograft humanized mouse model for evaluation of PD-1/PD-L1 blocker BMS202-induced immune responses in colorectal cancer.

The establishment of an in vivo mouse model mimicking human tumor-immune environments provides a promising platform for immunotherapy assessment, drug discovery and clinical decision guidance. To this end, we construct humanized NCG mice by transplanting human hCD34 + hematopoietic progenitors into non-obese diabetic (NOD) Cg- Prkdc scidIL2rg tm1Wjl /Sz (null; NCG) mice and monitoring the development of human hematopoietic and immune systems (Hu-NCG). The cell line-derived xenograft (CDX) Hu-NCG mouse models are set up to assess the outcome of immunotherapy mediated by the small molecule BMS202. As a PD-1/PD-L1 blocker, BMS202 shows satisfactory antitumour efficacy in the HCT116 and SW480 xenograft Hu-NCG mouse models. Mechanistically, BMS202 exerts antitumour efficacy by improving the tumor microenvironment and enhancing the infiltration of hCD8 + T cells and the release of hIFNγ in tumor tissue. Thus, tumor-bearing Hu-NCG mice are a suitable and important in vivo model for preclinical study, particularly in cancer immunotherapy.

Dry Matter Accumulation, Water Productivity and Quality of Potato in Response to Regulated Deficit Irrigation in a Desert Oasis Region.

As one of the most important food crops, the potato is widely planted in the oasis agricultural region of Northwest China. To ascertain the impact of regulated deficit irrigation (RDI) on various facets including dry matter accumulation, tuber yield, quality and water use efficiency (WUE) of potato plants, a two-growth season field experiment under mulched drip irrigation was conducted in the desert oasis region of Northwest China. Water deficits, applied at the seedling, tuber formation, tuber expansion and starch accumulation stages, encompassed two distinctive levels: mild (55-65% of field capacity, FC) and moderate (45-55% FC) deficit, with full irrigation (65-75% FC) throughout the growing season as the control (CK). The results showed that water deficit significantly reduced (p < 0.05) above-ground dry matter, water consumption and tuber yield compared to CK, and the reduction increased with the increasing water deficit. A mild water deficit at the tuber formation stage, without significantly reducing (p > 0.05) yield, could significantly increase WUE and irrigation water use efficiency (IWUE), with two-year average increases of 25.55% and 32.33%, respectively, compared to CK. Water deficit at the tuber formation stage increased starch content, whereas water deficit at tuber expansion stage significantly reduced starch, protein and reducing sugar content. Additionally, a comprehensive evaluation showed that a mild water deficit at the tuber formation stage is the optimal RDI strategy for potato production, providing a good balance between yield, quality and WUE. The results of this study can provide theoretical support for efficient and sustainable potato production in the desert oasis regions of Northwest China.

Unsupervised deep representation learning enables phenotype discovery for genetic association studies of brain imaging.

Understanding the genetic architecture of brain structure is challenging, partly due to difficulties in designing robust, non-biased descriptors of brain morphology. Until recently, brain measures for genome-wide association studies (GWAS) consisted of traditionally expert-defined or software-derived image-derived phenotypes (IDPs) that are often based on theoretical preconceptions or computed from limited amounts of data. Here, we present an approach to derive brain imaging phenotypes using unsupervised deep representation learning. We train a 3-D convolutional autoencoder model with reconstruction loss on 6130 UK Biobank (UKBB) participants' T1 or T2-FLAIR (T2) brain MRIs to create a 128-dimensional representation known as Unsupervised Deep learning derived Imaging Phenotypes (UDIPs). GWAS of these UDIPs in held-out UKBB subjects (n = 22,880 discovery and n = 12,359/11,265 replication cohorts for T1/T2) identified 9457 significant SNPs organized into 97 independent genetic loci of which 60 loci were replicated. Twenty-six loci were not reported in earlier T1 and T2 IDP-based UK Biobank GWAS. We developed a perturbation-based decoder interpretation approach to show that these loci are associated with UDIPs mapped to multiple relevant brain regions. Our results established unsupervised deep learning can derive robust, unbiased, heritable, and interpretable brain imaging phenotypes.

An atlas of genetic effects on the monocyte methylome across European and African populations.

Elucidating the genetic architecture of DNA methylation (DNAm) is crucial for decoding the etiology of complex diseases. However, current epigenomic studies often suffer from incomplete coverage of methylation sites and the use of tissues containing heterogeneous cell populations. To address these challenges, we present a comprehensive human methylome atlas based on deep whole-genome bisulfite sequencing (WGBS) and whole-genome sequencing (WGS) of purified monocytes from 298 European Americans (EA) and 160 African Americans (AA) in the Louisiana Osteoporosis Study. Our atlas enables the analysis of over 25 million DNAm sites. We identified 1,383,250 and 1,721,167 methylation quantitative trait loci (meQTLs) in cis -regions for EA and AA populations, respectively, with 880,108 sites shared between ancestries. While cis -meQTLs exhibited population-specific patterns, primarily due to differences in minor allele frequencies, shared cis -meQTLs showed high concordance across ancestries. Notably, cis -heritability estimates revealed significantly higher mean values in the AA population (0.09) compared to the EA population (0.04). Furthermore, we developed population-specific DNAm imputation models using Elastic Net, enabling methylome-wide association studies (MWAS) for 1,976,046 and 2,657,581 methylation sites in EA and AA, respectively. The performance of our MWAS models was validated through a systematic multi-ancestry analysis of 41 complex traits from the Million Veteran Program. Our findings bridge the gap between genomics and the monocyte methylome, uncovering novel methylation-phenotype associations and their transferability across diverse ancestries. The identified meQTLs, MWAS models, and data resources are freely available at www.gcbhub.org and https://osf.io/gct57/ .

Automated Type 2 Diabetes Case and Control Identification from the MIMIC-IV Database.

Phenotyping for Type 2 Diabetes (T2DM) is needed due to the increasing demand for T2DM research on electronic health records (EHRs). eMERGE is a reliable and interpretable rule-based algorithm for the identification of T2DM cases and controls in EHRs. MIMIC-IV, an extension of MIMIC-III, contains more than 520,000 hospital admissions and has become a valuable EHR database for secondary medical research. However, there was no prior work to extract T2DM cases and controls from MIMIC-IV, which requires a comprehensive knowledge of the database. Our work provided insight into the structure and data elements in MIMIC-IV and adapted eMERGE to accomplish the task. The results included MIMIC-IV's data tables and elements used, 12,735 cases and 9,828 controls of T2DM, and summary statistics of the cohorts in comparison with those on other EHR databases. They could be used for the development of statistical and machine learning models in future studies about the disease.

A PDA-Functionalized 3D Lung Scaffold Bioplatform to Construct Complicated Breast Tumor Microenvironment for Anticancer Drug Screening and Immunotherapy.

2D cell culture occupies an important place in cancer progression and drug discovery research. However, it limitedly models the "true biology" of tumors in vivo. 3D tumor culture systems can better mimic tumor characteristics for anticancer drug discovery but still maintain great challenges. Herein, polydopamine (PDA)-modified decellularized lung scaffolds are designed and can serve as a functional biosystem to study tumor progression and anticancer drug screening, as well as mimic the tumor microenvironment. PDA-modified scaffolds with strong hydrophilicity and excellent cell compatibility can promote cell growth and proliferation. After 96 h treatment with 5-FU, cisplatin, and DOX, higher survival rates in PDA-modified scaffolds are observed compared to nonmodified scaffolds and 2D systems. The E-cadhesion formation, HIF-1α-mediated senescence decrease, and tumor stemness enhancement can drive drug resistance and antitumor drug screening of breast cancer cells. Moreover, there is a higher survival rate of CD45+ /CD3+ /CD4+ /CD8+ T cells in PDA-modified scaffolds for potential cancer immunotherapy drug screening. This PDA-modified tumor bioplatform will supply some promising information for studying tumor progression, overcoming tumor resistance, and screening tumor immunotherapy drugs.

Design, synthesis, biological evaluation of urea substituted 1,2,5-oxadiazole-3-carboximidamides as novel indoleamine 2,3-dioxygenase-1 (IDO1) inhibitors.

Indoleamine 2,3-dioxygenase-1 (IDO1) has been considered as an attractive target for oncology immunotherapy due to its immunosuppressive effects on the tumor microenvironment. The most advanced IDO1 inhibitor epacadostat in combination with anti-PD-1 antibody failed to show desirable objective response. Epacadostat is now reevaluated in phase III clinical trials, but its pharmacokinetic (PK) properties are unsatisfactory. To further unravel the antitumor efficacy of IDO1 inhibitors, we designed a series of epacadostat analogues by introducing various urea-containing side chains. In particular, the most active compound 3 showed superior inhibitory potency against recombinant hIDO1 and hIDO1 in HeLa cells induced by interferon γ (IFNγ) relative to epacadostat (3, biochemical hIDO1 IC50 = 67.4 nM, HeLa hIDO1 IC50 = 17.6 nM; epacadostat, biochemical hIDO1 IC50 = 75.9 nM, HeLa hIDO1 IC50 = 20.6 nM). Moreover, compound 3 exhibited improved physicochemical properties and rat PK profile with better oral exposure and bioavailability compared with epacadostat. Importantly, this compound exhibited comparable antitumor efficacy with epacadostat in LLC syngeneic xenograft models. Hence, compound 3 represents a promising lead compound for discovery of more effective IDO1 inhibitors.

Recurrent neural network models (CovRNN) for predicting outcomes of patients with COVID-19 on admission to hospital: model development and validation using electronic health record data.

BACKGROUND: Predicting outcomes of patients with COVID-19 at an early stage is crucial for optimised clinical care and resource management, especially during a pandemic. Although multiple machine learning models have been proposed to address this issue, because of their requirements for extensive data preprocessing and feature engineering, they have not been validated or implemented outside of their original study site. Therefore, we aimed to develop accurate and transferrable predictive models of outcomes on hospital admission for patients with COVID-19. METHODS: In this study, we developed recurrent neural network-based models (CovRNN) to predict the outcomes of patients with COVID-19 by use of available electronic health record data on admission to hospital, without the need for specific feature selection or missing data imputation. CovRNN was designed to predict three outcomes: in-hospital mortality, need for mechanical ventilation, and prolonged hospital stay (>7 days). For in-hospital mortality and mechanical ventilation, CovRNN produced time-to-event risk scores (survival prediction; evaluated by the concordance index) and all-time risk scores (binary prediction; area under the receiver operating characteristic curve [AUROC] was the main metric); we only trained a binary classification model for prolonged hospital stay. For binary classification tasks, we compared CovRNN against traditional machine learning algorithms: logistic regression and light gradient boost machine. Our models were trained and validated on the heterogeneous, deidentified data of 247 960 patients with COVID-19 from 87 US health-care systems derived from the Cerner Real-World COVID-19 Q3 Dataset up to September 2020. We held out the data of 4175 patients from two hospitals for external validation. The remaining 243 785 patients from the 85 health systems were grouped into training (n=170 626), validation (n=24 378), and multi-hospital test (n=48 781) sets. Model performance was evaluated in the multi-hospital test set. The transferability of CovRNN was externally validated by use of deidentified data from 36 140 patients derived from the US-based Optum deidentified COVID-19 electronic health record dataset (version 1015; from January, 2007, to Oct 15, 2020). Exact dates of data extraction were masked by the databases to ensure patient data safety. FINDINGS: CovRNN binary models achieved AUROCs of 93·0% (95% CI 92·6-93·4) for the prediction of in-hospital mortality, 92·9% (92·6-93·2) for the prediction of mechanical ventilation, and 86·5% (86·2-86·9) for the prediction of a prolonged hospital stay, outperforming light gradient boost machine and logistic regression algorithms. External validation confirmed AUROCs in similar ranges (91·3-97·0% for in-hospital mortality prediction, 91·5-96·0% for the prediction of mechanical ventilation, and 81·0-88·3% for the prediction of prolonged hospital stay). For survival prediction, CovRNN achieved a concordance index of 86·0% (95% CI 85·1-86·9) for in-hospital mortality and 92·6% (92·2-93·0) for mechanical ventilation. INTERPRETATION: Trained on a large, heterogeneous, real-world dataset, our CovRNN models showed high prediction accuracy and transferability through consistently good performances on multiple external datasets. Our results show the feasibility of a COVID-19 predictive model that delivers high accuracy without the need for complex feature engineering. FUNDING: Cancer Prevention and Research Institute of Texas.

Discovery of Small-Molecule Inhibitors of the PD-1/PD-L1 Axis That Promote PD-L1 Internalization and Degradation.

Several monoclonal antibodies targeting the programmed cell death-1/programmed cell death-ligand 1 (PD-1/PD-L1) pathway have been used successfully in anticancer immunotherapy. Inherent limitations of antibody-based therapies remain, however, and alternative small-molecule inhibitors that can block the PD-1/PD-L1 axis are urgent needed. Herein, we report the discovery of compound 17 as a bifunctional inhibitor of PD-1/PD-L1 interactions. 17 inhibits PD-1/PD-L1 interactions and promotes dimerization, internalization, and degradation of PD-L1. 17 promotes cell-surface PD-L1 internalized into the cytosol and induces the degradation of PD-L1 in tumor cells through a lysosome-dependent pathway. Furthermore, 17 suppresses tumor growth in vivo by activating antitumor immunity. These results demonstrate that 17 targets the PD-1/PD-L1 axis and induces PD-L1 degradation.

Discovery of Small Molecules Simultaneously Targeting NAD(P)H:Quinone Oxidoreductase 1 and Nicotinamide Phosphoribosyltransferase: Treatment of Drug-Resistant Non-small-Cell Lung Cancer.

Targeting NAD+ metabolism has emerged as an effective anticancer strategy. Inspired by the synergistic antitumor effect between NAD(P)H:quinone oxidoreductase 1 (NQO1) substrates increasing the NAD consumption and nicotinamide phosphoribosyltransferase (NAMPT) inhibitors hampering the NAD synthesis, first-in-class small molecules simultaneously targeting NQO1 and NAMPT were identified through structure-based design. In particular, compound 10d is an excellent NQO1 substrate that is processed faster than TSA by NQO1 and exhibited a slightly decreased NAMPT inhibitory potency than that of FK866. It can selectively inhibit the proliferation of NQO1-overexpressing A549 cells and taxol-resistant A549/taxol cells and also induce cell apoptosis and inhibit cell migration in an NQO1- and NAMPT-dependent manner in A549/taxol cells. Significantly, compound 10d demonstrated excellent in vivo antitumor efficacy in the A549/taxol xenograft models with no significant toxicity. This proof-of-concept study affirms the feasibility of discovering small molecules that target NQO1 and NAMPT simultaneously, and it also provides a novel, effective, and selective anticancer strategy.

Phycocyanin-functionalized black phosphorus quantum dots enhance PDT/PTT therapy by inducing ROS and irreparable DNA damage.

To achieve synergistic photodynamic-photothermic therapy, we fabricate the novel phycocyanin (PC)-functionalized black phosphorus quantum dots (BPQDs) referred as PC@BPQDs through a one-step stirring method. PC@BPQDs are characterized by the feature of possessing both near-infrared (NIR) induced photothermal and photodynamic activity. The PC layer not only effectively alleviates plasma protein adsorption onto BPQDs, but also functionally boosts the photothermal therapy efficiency by enhanced ROS release, resulting in increased apoptosis in vitro. Moreover, PC@BPQDs eradicate tumors with high efficacy and low toxicity in vivo. Thus, PC@BPQDs have a promising potential in future therapeutic implications.

Rational design of a potent macrocyclic peptide inhibitor targeting the PD-1/PD-L1 protein-protein interaction.

We report optimization by rational design of JMPDP-027, a potent cyclic peptide that interferes with the PD-1/PD-L1 protein-protein interaction. JMPDP-027 shows a potent restoring ability towards T-cells with an EC50 of 5.9 nM that is comparable to that of the anti-PD-1 monoclonal antibody pembrolizumab. In addition, JMPDP-027 shows not only high resistance to enzymatic hydrolysis in human serum but also no observable toxicity and potent in vivo anticancer activity comparable to that of the mouse PD-L1 antibody in a colon carcinoma (CT26) model. Cyclic peptide antagonists of this sort may provide novel drug candidates for cancer immunotherapy.

The multiple regulation of metastasis suppressor NM23-H1 in cancer.

Metastasis is one of the leading causes of mortality in cancer patients. As the firstly identified metastasis suppressor, NM23-H1 has been endowed with expectation as a potent target in metastatic cancer therapy during the past decades. However, many challenges impede its clinical use. Accumulating evidence shows that NM23-H1 has a dichotomous role in tumor metastasis as a suppressor and promoter. It has potentially attributed to its versatile biochemical characteristics such as nucleoside diphosphate kinase (NDPK) activity, histidine kinase activity (HPK), exonuclease activity, and protein scaffold, which further augment the complexity and uncertainty of its physiological function. Simultaneously, tumor cells have evolved multiple ways to regulate the expression and function of NM23-H1 during tumorigenesis and metastasis. This review summarized and discussed the regulatory mechanisms of NM23-H1 in cancer including transcriptional activation, subcellular location, enzymatic activity, and protein degradation, which significantly modulate its anti-metastatic function.

Novel Biphenyl Pyridines as Potent Small-Molecule Inhibitors Targeting the Programmed Cell Death-1/Programmed Cell Death-Ligand 1 Interaction.

With the successful clinical application of anti-programmed cell death-1 (PD-1)/programmed cell death-ligand 1 (PD-L1) monoclonal antibodies (mAb), targeting the PD-1/PD-L1 interaction has become a promising method for the discovery of cancer therapy. Due to the inherent limitations of antibodies, it is necessary to search for small-molecule inhibitors against the PD-1/PD-L1 axis. We report the design, synthesis, and evaluation in vitro and in vivo of a series of novel biphenyl pyridines as the inhibitors of PD-1/PD-L1. 2-(((2-Methoxy-6-(2-methyl-[1,1'-biphenyl]-3-yl)pyridin-3-yl)methyl)amino)ethan-1-ol (24) was found to inhibit the PD-1/PD-L1 interaction with an IC50 value of 3.8 ± 0.3 nM and enhance the killing activity of tumor cells by immune cells. Compound 24 displays great pharmacokinetics (oral bioavailability of 22%) and significant in vivo antitumor activity in a CT26 mouse model. Flow cytometry and immunohistochemistry data indicated that compound 24 activates the immune activity in tumors. These results suggest that compound 24 is a promising small-molecule inhibitor against the PD-1/PD-L1 axis and merits further development.

Dual nicotinamide phosphoribosyltransferase and epidermal growth factor receptor inhibitors for the treatment of cancer.

Multitarget drugs have emerged as a promising treatment modality in modern anticancer therapy. Taking advantage of the synergy of NAMPT and EGFR inhibition, we have developed the first compounds that serve as dual inhibitors of NAMPT and EGFR. On the basis of CHS828 and erlotinib, a series of hybrid molecules were successfully designed and synthesized by merging of the pharmacophores. Among the compounds that were synthesized, compound 28 showed good NAMPT and EGFR inhibition, and excellent in vitro anti-proliferative activity. Compound 28, which is a new chemotype devoid of a Michael receptor, strongly inhibited the proliferation of several cancer cell lines, including H1975 non-small cell lung cancer cells harboring the EGFRL858R/T790M mutation. More importantly, it imparted significant in vivo antitumor efficacy in a human NSCLC (H1975) xenograft nude mouse model. This study provides promising leads for the development of novel antitumor agents and valuable pharmacological probes for the assessment of dual inhibition in NAMPT and EGFR pathway with a single inhibitor.

Frequency Characteristic of Resonant Micro Fluidic Chip for Oil Detection Based on Resistance Parameter.

Monitoring the working condition of hydraulic equipment is significance in industrial fields. The abnormal wear of the hydraulic system can be revealed by detecting the variety and size of micro metal debris in the hydraulic oil. We thus present the design and implementation of a micro detection system of hydraulic oil metal debris based on inductor capacitor (LC) resonant circuit in this paper. By changing the resonant frequency of the micro fluidic chip, we can detect the metal debris of hydraulic oil and analyze the sensitivity of the micro fluidic chip at different resonant frequencies. We then obtained the most suitable resonant frequency. The chip would generate a positive resistance pulse when the iron particles pass through the detection area and the sensitivity of the chip decreased with resonant frequency. The chip would generate a negative resistance pulse when the copper particles pass through the detection area and the sensitivity of the chip increased with resonant frequency. The experimental results show that the change of resonant frequency has a great effect on the copper particles and little on the iron particles. Thus, a relatively big resonant frequency can be selected for chip designing and testing. In practice, we can choose a relatively big resonant frequency in this micro fluidic chip designing. The resonant micro fluidic chip is capable of detecting 20⁻30 µm iron particles and 70⁻80 µm copper particles at 0.9 MHz resonant frequency.