Development and Validation of an Explainable Deep Learning Model to Predict In-Hospital Mortality for Patients With Acute Myocardial Infarction: Algorithm Development and Validation Study.

BACKGROUND: Acute myocardial infarction (AMI) is one of the most severe cardiovascular diseases and is associated with a high risk of in-hospital mortality. However, the current deep learning models for in-hospital mortality prediction lack interpretability. OBJECTIVE: This study aims to establish an explainable deep learning model to provide individualized in-hospital mortality prediction and risk factor assessment for patients with AMI. METHODS: In this retrospective multicenter study, we used data for consecutive patients hospitalized with AMI from the Chongqing University Central Hospital between July 2016 and December 2022 and the Electronic Intensive Care Unit Collaborative Research Database. These patients were randomly divided into training (7668/10,955, 70%) and internal test (3287/10,955, 30%) data sets. In addition, data of patients with AMI from the Medical Information Mart for Intensive Care database were used for external validation. Deep learning models were used to predict in-hospital mortality in patients with AMI, and they were compared with linear and tree-based models. The Shapley Additive Explanations method was used to explain the model with the highest area under the receiver operating characteristic curve in both the internal test and external validation data sets to quantify and visualize the features that drive predictions. RESULTS: A total of 10,955 patients with AMI who were admitted to Chongqing University Central Hospital or included in the Electronic Intensive Care Unit Collaborative Research Database were randomly divided into a training data set of 7668 (70%) patients and an internal test data set of 3287 (30%) patients. A total of 9355 patients from the Medical Information Mart for Intensive Care database were included for independent external validation. In-hospital mortality occurred in 8.74% (670/7668), 8.73% (287/3287), and 9.12% (853/9355) of the patients in the training, internal test, and external validation cohorts, respectively. The Self-Attention and Intersample Attention Transformer model performed best in both the internal test data set and the external validation data set among the 9 prediction models, with the highest area under the receiver operating characteristic curve of 0.86 (95% CI 0.84-0.88) and 0.85 (95% CI 0.84-0.87), respectively. Older age, high heart rate, and low body temperature were the 3 most important predictors of increased mortality, according to the explanations of the Self-Attention and Intersample Attention Transformer model. CONCLUSIONS: The explainable deep learning model that we developed could provide estimates of mortality and visual contribution of the features to the prediction for a patient with AMI. The explanations suggested that older age, unstable vital signs, and metabolic disorders may increase the risk of mortality in patients with AMI.

Multi-Responsive Peptide-Based Ultrathin Nanosheets Prepared by a Horizontal Monolayer Assembly.

In this study, peptide-based self-assembled nanosheets with a thickness of approximately 1 nm were prepared using a hierarchical covalent physical fabrication strategy. The covalent alternating polymerization of helical peptide E3 with an azobenzene (AZO) structure yielded copolymers CoP(E3-AZO), which physically self-assembled into ultrathin nanosheets in an unanticipated two-dimensional horizontal monolayer arrangement. This special monolayer arrangement enabled the thickness of the nanosheets to be equal to the cross-sectional diameter of a single linear copolymer, which is a rare phenomenon. Molecular dynamics simulations suggested that the synergistic effect of multiple molecular interactions drives the self-assembly of CoP(E3-AZO) into nanosheets and that various methods, including phototreatment, pH adjustment, the addition of additives, and introduction of cosolvents, can alter the molecular interactions and modulate the self-assembly of CoP(E3-AZO), yielding diverse nanostructures. Remarkably, the ultrathin nanosheets selectively inhibited cancer cells at certain concentrations.

Annotations of four high-quality indigenous chicken genomes identify more than one thousand missing genes in subtelomeric regions and micro-chromosomes with high G/C contents.

BACKGROUND: Although multiple chicken genomes have been assembled and annotated, the numbers of protein-coding genes in chicken genomes and their variation among breeds are still uncertain due to the low quality of these genome assemblies and limited resources used in their gene annotations. To fill these gaps, we recently assembled genomes of four indigenous chicken breeds with distinct traits at chromosome-level. In this study, we annotated genes in each of these assembled genomes using a combination of RNA-seq- and homology-based approaches. RESULTS: We identified varying numbers (17,497-17,718) of protein-coding genes in the four indigenous chicken genomes, while recovering 51 of the 274 "missing" genes in birds in general, and 36 of the 174 "missing" genes in chickens in particular. Intriguingly, based on deeply sequenced RNA-seq data collected in multiple tissues in the four breeds, we found 571 ~ 627 protein-coding genes in each genome, which were missing in the annotations of the reference chicken genomes (GRCg6a and GRCg7b/w). After removing redundancy, we ended up with a total of 1,420 newly annotated genes (NAGs). The NAGs tend to be found in subtelomeric regions of macro-chromosomes (chr1 to chr5, plus chrZ) and middle chromosomes (chr6 to chr13, plus chrW), as well as in micro-chromosomes (chr14 to chr39) and unplaced contigs, where G/C contents are high. Moreover, the NAGs have elevated quadruplexes G frequencies, while both G/C contents and quadruplexes G frequencies in their surrounding regions are also high. The NAGs showed tissue-specific expression, and we were able to verify 39 (92.9%) of 42 randomly selected ones in various tissues of the four chicken breeds using RT-qPCR experiments. Most of the NAGs were also encoded in the reference chicken genomes, thus, these genomes might harbor more genes than previously thought. CONCLUSION: The NAGs are widely distributed in wild, indigenous and commercial chickens, and they might play critical roles in chicken physiology. Counting these new genes, chicken genomes harbor more genes than originally thought.

NextDenovo: an efficient error correction and accurate assembly tool for noisy long reads.

Long-read sequencing data, particularly those derived from the Oxford Nanopore sequencing platform, tend to exhibit high error rates. Here, we present NextDenovo, an efficient error correction and assembly tool for noisy long reads, which achieves a high level of accuracy in genome assembly. We apply NextDenovo to assemble 35 diverse human genomes from around the world using Nanopore long-read data. These genomes allow us to identify the landscape of segmental duplication and gene copy number variation in modern human populations. The use of NextDenovo should pave the way for population-scale long-read assembly using Nanopore long-read data.

Quercetin inhibition of porcine intestinal alpha coronavirus in vitro and in vivo.

BACKGROUND: Porcine acute diarrhea syndrome coronavirus (SADS-CoV) is one of the novel pathogens responsible for piglet diarrhea, contributing to substantial economic losses in the farming sector. The broad host range of SADS-CoV raises concerns regarding its potential for cross-species transmission. Currently, there are no effective means of preventing or treating SADS-CoV infection, underscoring the urgent need for identifying efficient antiviral drugs. This study focuses on evaluating quercetin as an antiviral agent against SADS-CoV. RESULTS: In vitro experiments showed that quercetin inhibited SADS-CoV proliferation in a concentration-dependent manner, targeting the adsorption and replication stages of the viral life cycle. Furthermore, quercetin disrupts the regulation of the P53 gene by the virus and inhibits host cell cycle progression induced by SADS-CoV infection. In vivo experiments revealed that quercetin effectively alleviated the clinical symptoms and intestinal pathological damage caused by SADS-CoV-infected piglets, leading to reduced expression levels of inflammatory factors such as TLR3, IL-6, IL-8, and TNF-α. CONCLUSIONS: Therefore, this study provides compelling evidence that quercetin has great potential and promising applications for anti- SADS-CoV action.

Resveratrol alleviates acute lung injury in mice by promoting Pink1/Parkin-related mitophagy and inhibiting NLRP3 inflammasome activation.

Acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) are characterized by rapid onset and widespread inflammation in the lungs, often leading to respiratory failure. These conditions can be triggered by various factors, resulting in a severe inflammatory response within the lungs. Resveratrol, a polyphenolic compound found in grapes and peanuts, is renowned for its potent antioxidative and anti-inflammatory properties. In this study, we investigated how resveratrol protects against lipopolysaccharide (LPS)-induced ALI in mice. We established mouse models of LPS-induced ALI and inflammation in bronchoalveolar lavage fluid (BALF) macrophages. Through histopathological examination, immunofluorescence, western blot, enzyme-linked immunosorbent assay (ELISA), and transmission electron microscopy (TEM), we assessed the impact of resveratrol on the activation of NOD-like receptor thermal protein domain-associated protein 3 (NLRP3) inflammasomes and the process of mitophagy. Our findings indicate that resveratrol significantly mitigated the lung injury and inflammation caused by LPS. This was achieved by inhibiting the oligomerization of apoptosis-associated speck-like protein containing a caspase recruitment domain (ASC) and the activation of NLRP3 inflammasomes. Resveratrol also reduced the levels of IL-1ß and IL-18 in serum and BALF, decreased caspase-1 expression, and diminished macrophage pyroptosis. Furthermore, it upregulated Pink1, Parkin, Beclin-1, Autophagy-Related 5 (Atg5), and Microtubule-Associated Proteins 1 A/1B Light Chain 3B (LC3B-II), thereby enhancing mitophagy. Conversely, mitophagy was inhibited by Pink1 siRNA. In conclusion, resveratrol ameliorated ALI in mice, potentially by inhibiting the activation of NLRP3 inflammasomes, activating the Pink1/Parkin pathway, and promoting mitophagy.

Artificial selection footprints in indigenous and commercial chicken genomes.

BACKGROUND: Although many studies have been done to reveal artificial selection signatures in commercial and indigenous chickens, a limited number of genes have been linked to specific traits. To identify more trait-related artificial selection signatures and genes, we re-sequenced a total of 85 individuals of five indigenous chicken breeds with distinct traits from Yunnan Province, China. RESULTS: We found 30 million non-redundant single nucleotide variants and small indels (< 50 bp) in the indigenous chickens, of which 10 million were not seen in 60 broilers, 56 layers and 35 red jungle fowls (RJFs) that we compared with. The variants in each breed are enriched in non-coding regions, while those in coding regions are largely tolerant, suggesting that most variants might affect cis-regulatory sequences. Based on 27 million bi-allelic single nucleotide polymorphisms identified in the chickens, we found numerous selective sweeps and affected genes in each indigenous chicken breed and substantially larger numbers of selective sweeps and affected genes in the broilers and layers than previously reported using a rigorous statistical model. Consistent with the locations of the variants, the vast majority (~ 98.3%) of the identified selective sweeps overlap known quantitative trait loci (QTLs). Meanwhile, 74.2% known QTLs overlap our identified selective sweeps. We confirmed most of previously identified trait-related genes and identified many novel ones, some of which might be related to body size and high egg production traits. Using RT-qPCR, we validated differential expression of eight genes (GHR, GHRHR, IGF2BP1, OVALX, ELF2, MGARP, NOCT, SLC25A15) that might be related to body size and high egg production traits in relevant tissues of relevant breeds. CONCLUSION: We identify 30 million single nucleotide variants and small indels in the five indigenous chicken breeds, 10 million of which are novel. We predict substantially more selective sweeps and affected genes than previously reported in both indigenous and commercial breeds. These variants and affected genes are good candidates for further experimental investigations of genotype-phenotype relationships and practical applications in chicken breeding programs.

Predictive value of NT-proBNP and hs-TnT for outcomes after pediatric congenital cardiac surgery.

BACKGROUND: The available evidence regarding the predictive value of troponins and natriuretic peptides for early postoperative outcomes in pediatrics is limited, controversial, and based on small sample sizes. We aimed to investigate the association of N-terminal pro B-type natriuretic peptide (NT-proBNP) and high-sensitivity troponin T (hs-TnT) with the in-hospital adverse outcomes after congenital cardiac surgeries. METHODS: A secondary analysis based on a prospective study of pediatric congenital heart disease (CHD) patients was conducted to investigate the association of NT-proBNP and hs-TnT tested within 6 hours postoperatively with in-hospital adverse events. A multivariate logistic regression analysis with a minimum P value approach was used to identify the optimal thresholds of NT-proBNP and hs-TnT for risk stratification. RESULTS: NT-proBNP and hs-TnT are positively correlated with cardiopulmonary bypass time, mechanical ventilation duration, and pediatric intensive care unit stay. The predictive performance of NT-proBNP is excellent for adverse events in both patients younger than 1 year [area under the curve (AUC): 0.771, 0.693-0.850] and those older than 1 year (AUC: 0.839, 0.757-0.922). However, hs-TnT exhibited a satisfactory predictive value solely in patients aged over 1 year. (AUC: 0.784, 0.717-0.852). NT-proBNP levels of 2000 to 10000 ng/L [Odds Ratio (OR): 3.79, 1.47-9.76) and exceeding 10000 ng/L (OR: 12.21, 3.66-40.80) were associated with a higher risk of postoperative adverse events in patients younger than 1 year. Patients older than 1 year, with NT-proBNP higher than 500 ng/L (OR: 15.09, 6.05-37.66) or hs-TnT greater than 1200 ng/L (OR: 5.50, 1.47-20.59), had a higher incidence of postoperative adverse events. CONCLUSIONS: NT-proBNP and hs-TnT tested within postoperative 6 hours demonstrated significant predictive value for postoperative adverse events in CHD patients older than 1 year. However, among CHD patients younger than 1 year, only NT-proBNP exhibited commendable predictive performance for postoperative adverse events.

The discovery of potent USP2/USP8 dual-target inhibitors for the treatment of breast cancer via structure guided optimization of ML364.

USP2 and USP8 are crucial in the development and progression of breast cancer, primarily through the stabilization of protein substrates such as Her2 and ERα. The dual-target inhibitor ML364, targeting both USP2 and USP8, has garnered significant interest in recent research. In this study, we developed a series of ML364 derivatives using ligand-based drug design strategies. The standout compound, LLK203, demonstrated enhanced inhibitory activity, showing a 4-fold increase against USP2 and a 9-fold increase against USP8, compared to the parent molecule. In MCF-7 breast cancer cells, LLK203 effectively degraded key proteins involved in cancer progression and notably inhibited cell proliferation. Moreover, LLK203 exhibited potent in vivo efficacy in the 4T1 homograft model, while maintaining a low toxicity profile. These results underscore the potential of LLK203 as a promising dual-target inhibitor of USP2/USP8 for breast cancer treatment.

High quality assemblies of four indigenous chicken genomes and related functional data resources.

Many lines of evidence indicate that red jungle fowl (RJF) is the primary ancestor of domestic chickens. Although multiple versions of RJF (galgal2-galgal5 and GRCg6a) and commercial chickens (GRCg7b/w and Huxu) genomes have been assembled since 2004, no high-quality indigenous chicken genomes have been assembled, hampering the understanding of chicken domestication and evolution. To fill the gap, we sequenced the genomes of four indigenous chickens with distinct morphological traits in southwest China, using a combination of short, long and Hi-C reads. We assembled each genome (~1.0 Gb) into 42 chromosomes with chromosome N50 90.5-90.9 Mb, amongst the highest quality of chicken genome assemblies. To provide resources for gene annotation and functional analysis, we also sequenced transcriptomes of 10 tissues for each of the four chickens. Moreover, we corrected many mis-assemblies and assembled missing micro-chromosomes 29 and 34-39 for GRCg6a. Our assemblies, sequencing data and the correction of GRCg6a can be valuable resources for studying chicken domestication and evolution.

The potential role of hydrogen sulfide in cancer cell apoptosis.

For a long time, hydrogen sulfide (H2S) has been considered a toxic compound, but recent studies have found that H2S is the third gaseous signaling molecule which plays a vital role in physiological and pathological conditions. Currently, a large number of studies have shown that H2S mediates apoptosis through multiple signaling pathways to participate in cancer occurrence and development, for example, PI3K/Akt/mTOR and MAPK signaling pathways. Therefore, the regulation of the production and metabolism of H2S to mediate the apoptotic process of cancer cells may improve the effectiveness of cancer treatment. In this review, the role and mechanism of H2S in cancer cell apoptosis in mammals are summarized.

High-quality genome assembly of a C. crossoptilon and related functional and genetics data resources.

There are four species in the Crossoptilon genus inhibiting at from very low to very high altitudes across China, and they are in varying levels of danger of extinction. To better understand the genetic basis of adaptation to high altitudes and genetic changes due to bottleneck, we assembled the genome (~1.02 Gb) of a white eared pheasant (WT) (Crossoptilon crossoptilon) inhibiting at high altitudes (3,000~7,000 m) in northwest of Yunnan province, China, using a combination of Illumina short reads, PacBio long reads and Hi-C reads, with a contig N50 of 19.63 Mb and only six gaps. To further provide resources for gene annotation as well as functional and population genetics analyses, we sequenced transcriptomes of 20 major tissues of the WT individual and re-sequenced another 10 WT individuals and a blue eared pheasant (Crossoptilon auritum) individual inhabiting at intermediate altitudes (1,500~3,000 m). Our assembled WT genome, transcriptome data, and DNA sequencing data can be valuable resources for studying the biology, evolution and developing conservation strategies of these endangered species.

Divergent Evolutionary Rates of Primate Brain Regions as Revealed by Genomics and Transcriptomics.

Although the primate brain contains numerous functionally distinct structures that have experienced diverse genetic changes during the course of evolution and development, these changes remain to be explored in detail. Here we utilize two classic metrics from evolutionary biology, the evolutionary rate index (ERI) and the transcriptome age index (TAI), to investigate the evolutionary alterations that have occurred in each area and developmental stage of the primate brain. We observed a higher evolutionary rate for those genes expressed in the non-cortical areas during primate evolution, particularly in human, with the highest rate of evolution being exhibited at brain developmental stages between late infancy and early childhood. Further, the transcriptome age of the non-cortical areas was lower than that of the cerebral cortex, with the youngest age apparent at brain developmental stages between late infancy and early childhood. Our exploration of the evolutionary patterns manifest in each brain area and developmental stage provides important reference points for further research into primate brain evolution.

Deciphering the Role of Rapidly Evolving Conserved Elements in Primate Brain Development and Exploring Their Potential Involvement in Alzheimer's Disease.

Although previous studies have identified human-specific accelerated regions as playing a key role in the recent evolution of the human brain, the characteristics and cellular functions of rapidly evolving conserved elements (RECEs) in ancestral primate lineages remain largely unexplored. Here, based on large-scale primate genome assemblies, we identify 888 RECEs that have been highly conserved in primates that exhibit significantly accelerated substitution rates in the ancestor of the Simiiformes. This primate lineage exhibits remarkable morphological innovations, including an expanded brain mass. Integrative multiomic analyses reveal that RECEs harbor sequences with potential cis-regulatory functions that are activated in the adult human brain. Importantly, genes linked to RECEs exhibit pronounced expression trajectories in the adult brain relative to the fetal stage. Furthermore, we observed an increase in the chromatin accessibility of RECEs in oligodendrocytes from individuals with Alzheimer's disease (AD) compared to that of a control group, indicating that these RECEs may contribute to brain aging and AD. Our findings serve to expand our knowledge of the genetic underpinnings of brain function during primate evolution.

Tumor Necrosis Factor Alpha-Induced Protein-3 Inhibits the Activation of Hepatic Stellate Cells by Regulating the p65 Molecule in the NF-κB Signaling Pathway.

Tumor necrosis factor alpha-induced protein-3, also called A20, is a zinc-finger protein that participates in various inflammatory responses; however, the putative relationship between A20 and hepatic fibrosis remains unelucidated. Therefore, we investigated the role and mechanism of action of A20 in activating hepatic stellate cells (HSC) during the progression of hepatic fibrosis. Cell counting kit-8 (CCK8), colony growth, transwell assays, cell cycle analysis, and apoptosis assays were performed to explore the effect of A20 on cell function in vitro. An interspecies intravenous injection of the adeno-associated virus was used to assess the in vivo role of A20. The regulation of A20 on p65 was detected using mass spectrometry and immunoprecipitation. Our findings revealed that A20 was highly expressed in the liver tissues of patients with hepatic fibrosis and that the expression level of A20 in the liver tissue was closely correlated with the stage of liver fibrosis. In the LX-2 cell line, the downregulation of A20 upregulated the expression of fibrosis-related proteins and increased the expression of inflammatory factors, indicating the activation of HSC and vice versa. In addition, overexpression of A20 in mice reduced the degree of liver fibrosis in thioacetamide model mice. Finally, co-immunoprecipitation demonstrated that A20 could interact with p65. Hence, A20 inhibits HSC activation by binding to p65.

Printed Divisional Optical Biochip for Multiplex Visualizable Exosome Analysis at Point-of-Care.

Rapid detection of various exosomes is of great significance in early diagnosis and postoperative monitoring of cancers. Here, a divisional optical biochip is reported for multiplex exosome analysis via combining the self-assembly of nanochains and precise surface patterning. Arising from resonance-induced near-field enhancement, the nanochains show distinct color changes after capturing target exosomes for direct visual detection. Then, a series of divisional nanochain-based biochips conjugated with several specific antibodies are fabricated through designed hydrophilic and hydrophobic patterns. Because of the significant wettability difference, one sample droplet is precisely self-splitting into several microdroplets enabling simultaneous identification of multiple target exosomes in 30 min with a sensitivity of 6 × 107 particles mL-1 , which is about two orders lower than enzyme-linked immunosorbent assay. Apart from the trace amount detection, excellent semiquantitative capability is demonstrated to distinguish clinical exosomes from glioblastoma patients and healthy people. This method is simple, versatile, and highly efficient that can be extended as a diagnostic tool for many diseases, promoting the development of liquid biopsy.

A human-specific insertion promotes cell proliferation and migration by enhancing TBC1D8B expression.

Human-specific insertions play important roles in human phenotypes and diseases. Here we reported a 446-bp insertion (Insert-446) in intron 11 of the TBC1D8B gene, located on chromosome X, and traced its origin to a portion of intron 6 of the EBF1 gene on chromosome 5. Interestingly, Insert-446 was present in the human Neanderthal and Denisovans genomes, and was fixed in humans after human-chimpanzee divergence. We have demonstrated that Insert-446 acts as an enhancer through binding transcript factors that promotes a higher expression of human TBC1D8B gene as compared with orthologs in macaques. In addition, over-expression TBC1D8B promoted cell proliferation and migration through "a dual finger" catalytic mechanism (Arg538 and Gln573) in the TBC domain in vitro and knockdown of TBC1D8B attenuated tumorigenesis in vivo. Knockout of Insert-446 prevented cell proliferation and migration in cancer and normal cells. Our results reveal that the human-specific Insert-446 promotes cell proliferation and migration by upregulating the expression of TBC1D8B gene. These findings provide a significant insight into the effects of human-specific insertions on evolution.

Economic Evaluation of Neoadjuvant Versus Adjuvant Chemotherapy in Cancer Treatment: A Systematic Review and Meta-Analysis.

OBJECTIVES: In the absence of evidence on whether neoadjuvant (NAC) or adjuvant chemotherapy (AC) is more beneficial for various tumor treatments, economic evaluation (EE) can assist medical decision making. There is limited evidence on their cost-effectiveness and their prospective evaluation is less likely in the future. Therefore, a systematic review and meta-analysis about EE for NAC versus AC in solid tumor help compare these therapies from various perspectives. METHODS: Various databases were searched for studies published from inception to 2021. This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses reporting guidelines and economic-specific guidelines. The data were pooled using a random effects model when possible. RESULTS: The retrieval identified 15 EE studies of NAC versus AC in 8 types of cancer. NAC is the dominant strategy for pancreatic, head and neck, rectal, prostate cancers and colorectal liver metastases. For ovarian cancer, NAC is cost-effective with a lower cost and higher or similar quality-adjusted life-year. There were no significant differences in cost and outcomes for lung cancer. For stage IV or high-risk patients with ovarian or prostate cancer, NAC was cost-effective but not for patients who were not high risk. CONCLUSIONS: The EEs results for NAC versus AC were inconsistent because of their different model structures, assumptions, cost inclusions, and a shortage of studies. There are multiple sources of heterogeneity across EEs evidence synthesis. More high-quality EE studies on NAC versus AC in initial cancer treatment are necessary.