Instance-Specific Loss-Weighted Decoding for Decomposition-Based Multiclass Classification.

Multiclass classification problems are often addressed by decomposing them into a set of binary classification tasks. A critical step in this approach is the effective aggregation of predictions from each decomposed binary classifier to yield the final multiclass prediction, a process known as decoding. Existing studies have ignored the varying generalization ability of each binary classifier across different samples during decoding, potentially leading to suboptimal performance. In this article, we propose an instance-specific loss-weighted (ILW) decoding strategy that gauges the generalization ability of each binary classifier for one specific sample based on its neighboring samples. This estimated generalization ability is then used to adjust the importance of the binary classifier in determining the sample's final prediction. Experimental results validate the effectiveness of the ILW decoding strategy. Furthermore, we demonstrate that softmax regression can be reinterpreted as a one-versus-rest (OvR) decomposition-based multiclass classification algorithm, enabling the application of our decoding strategy to enhance its performance. Comparative studies clearly demonstrate the superiority of the improved softmax regression over its traditional counterpart.

Variational Label Enhancement for Instance-Dependent Partial Label Learning.

Partial label learning (PLL) is a form of weakly supervised learning, where each training example is linked to a set of candidate labels, among which only one label is correct. Most existing PLL approaches assume that the incorrect labels in each training example are randomly picked as the candidate labels. However, in practice, this assumption may not hold true, as the candidate labels are often instance-dependent. In this paper, we address the instance-dependent PLL problem and assume that each example is associated with a latent label distribution where the incorrect label with a high degree is more likely to be annotated as a candidate label. Motivated by this consideration, we propose two methods VALEN and MILEN, which train the predictive model via utilizing the latent label distributions recovered by the label enhancement process. Specifically, VALEN recovers the latent label distributions via inferring the variational posterior density parameterized by an inference model with the deduced evidence lower bound. MILEN recovers the latent label distribution by adopting the variational approximation to bound the mutual information among the latent label distribution, observed labels and augmented instances. Experiments on benchmark and real-world datasets validate the effectiveness of the proposed methods.

Case report: Cytokine and miRNA profiling in multisystem inflammatory syndrome in children.

Multisystem inflammatory syndrome in children (MIS-C) is an imperative pediatric inflammatory condition closely linked to COVID-19, which garners substantial attention since the onset of the pandemic. Like Kawasaki illness, this condition is characterized by an overactive immune response, leading to symptoms including pyrexia, cardiac and renal complications. To elucidate the pathogenesis of MIS-C and identify potential biomarkers, we conducted an extensive examination of specific cytokines (IL-6, IL-1ß, IL-6R, IL-10, and TNF-α) and microRNA (miRNA) expression profiles at various intervals (ranging from 3 to 20 days) in the peripheral blood sample of a severely affected MIS-C patient. Our investigation revealed a gradual decline in circulating levels of IL-6, IL-1ß, IL-10, and TNF-α following intravenous immune globulin (IVIG) therapy. Notably, IL-6 exhibited a significant reduction from 74.30 to 1.49 pg./mL, while IL-6R levels remained consistently stable throughout the disease course. Furthermore, we observed an inverse correlation between the expression of hsa-miR-596 and hsa-miR-224-5p and the aforementioned cytokines. Our findings underscore a robust association between blood cytokine and miRNA concentrations and the severity of MIS-C. These insights enhance our understanding of the genetic regulatory mechanisms implicated in MIS-C pathogenesis, offering potential avenues for early biomarker detection and therapy monitoring through miRNA analysis.

Label-Specific Time-Frequency Energy-Based Neural Network for Instrument Recognition.

Predominant instrument recognition plays a vital role in music information retrieval. This task involves identifying and categorizing the dominant instruments present in a piece of music based on their distinctive time-frequency characteristics and harmonic distribution. Existing predominant instrument recognition approaches mainly focus on learning implicit mappings (such as deep neural networks) from time-domain or frequency-domain representations of music audio to instrument labels. However, different instruments playing in polyphonic music produce local superposed time-frequency representations while most implicit models could be sensitive to such local data changes. This thus poses a challenge for these implicit methods to accurately capture the unique harmonic features of each instrument. To address this challenge, considering that the complete harmonic information of an instrument is also distributed across a wide range of frequencies, we design a label-specific time-frequency feature learning approach to convert the task of building implicit classification mappings into the process of extracting and matching features that are specific to each instrument, as a result, a new explicit learning model: label-specific time-frequency energy-based neural network (LSTN) is proposed. Unlike existing implicit models, LSTN not only extracts their commonly used local time-frequency features but also incorporates time-domain factors and frequency-domain factors in its energy function to explicitly parameterize the long-term correlation and long-frequency correlation features. Using the extracted time-frequency features and the two long correlation features as instrument label-specific features, LSTN detects whether the harmonic distribution of each instrument appears in polyphonic music on both long time-frequency scales and local time-frequency scales to mitigate the challenges posed by local superposed representations. We conduct an analysis of the complexity and the convergence of LSTN, then experiments conducted on benchmark datasets demonstrate the superiority of LSTN over other established instrument recognition algorithms.

CRISPR-Cas9-mediated construction of a cotton CDPK mutant library for identification of insect-resistance genes.

Calcium-dependent protein kinases (CDPKs) act as key signal transduction enzymes in plants, especially in response to diverse stresses, including herbivory. In this study, a comprehensive analysis of the CDPK gene family in upland cotton revealed that GhCPKs are widely expressed in multiple cotton tissues and respond positively to various biotic and abiotic stresses. We developed a strategy for screening insect-resistance genes from a CRISPR-Cas9 mutant library of GhCPKs. The library was created using 246 single-guide RNAs targeting the GhCPK gene family to generate 518 independent T0 plants. The average target-gene coverage was 86.18%, the genome editing rate was 89.49%, and the editing heritability was 82%. An insect bioassay in the field led to identification of 14 GhCPK mutants that are resistant or susceptible to insects. The mutant that showed the clearest insect resistance, cpk33/74 (in which the homologous genes GhCPK33 and GhCPK74 were knocked out), was selected for further study. Oral secretions from Spodoptera litura induced a rapid influx of Ca2+ in cpk33/74 leaves, resulting in a significant increase in jasmonic acid content. S-adenosylmethionine synthase is an important protein involved in plant stress response, and protein interaction experiments provided evidence for interactions of GhCPK33 and GhCPK74 with GhSAMS1 and GhSAM2. In addition, virus-induced gene silencing of GhSAMS1 and GhSAM2 in cotton impaired defense against S. litura. This study demonstrates an effective strategy for constructing a mutant library of a gene family in a polyploid plant species and offers valuable insights into the role of CDPKs in the interaction between plants and herbivorous insects.

Radiogenomic analysis for predicting lymph node metastasis and molecular annotation of radiomic features in pancreatic cancer.

BACKGROUND: To provide a preoperative prediction model for lymph node metastasis in pancreatic cancer patients and provide molecular information of key radiomic features. METHODS: Two cohorts comprising 151 and 54 pancreatic cancer patients were included in the analysis. Radiomic features from the tumor region of interests were extracted by using PyRadiomics software. We used a framework that incorporated 10 machine learning algorithms and generated 77 combinations to construct radiomics-based models for lymph node metastasis prediction. Weighted gene coexpression network analysis (WGCNA) was subsequently performed to determine the relationships between gene expression levels and radiomic features. Molecular pathways enrichment analysis was performed to uncover the underlying molecular features. RESULTS: Patients in the in-house cohort (mean age, 61.3 years ± 9.6 [SD]; 91 men [60%]) were separated into training (n = 105, 70%) and validation (n = 46, 30%) cohorts. A total of 1,239 features were extracted and subjected to machine learning algorithms. The 77 radiomic models showed moderate performance for predicting lymph node metastasis, and the combination of the StepGBM and Enet algorithms had the best performance in the training (AUC = 0.84, 95% CI = 0.77-0.91) and validation (AUC = 0.85, 95% CI = 0.73-0.98) cohorts. We determined that 15 features were core variables for lymph node metastasis. Proliferation-related processes may respond to the main molecular alterations underlying these features. CONCLUSIONS: Machine learning-based radiomics could predict the status of lymph node metastasis in pancreatic cancer, which is associated with proliferation-related alterations.

Enhancing Cordycepin in Functional Porridge through Rice Varietal Fermentation with Cordyceps militaris (Ascomycetes) and Utilizing Non-Targeted Metabolomics for Process Enhancement.

Cordyceps militaris, a medicinal fungus rich in cordycepin, shows promise in treating diseases such as cancer, respiratory issues, and COVID-19. This study examines the impact of different Taiwanese rice varieties on its solid-state fermentation, focusing on optimizing cordycepin production. The results indicated that the cordycepin yield was indeed affected by the type of rice used. In terms of the fruiting bodies, germ rice resulted in the highest yield (13.1 ± 0.36 mg/g), followed by brown rice (11.9 ± 0.26 mg/g). In the rice culture medium (RCM), brown rice led to the highest yield (4.77 ± 0.06 mg/g). Using gas chromatography-mass spectrometry and untargeted metabolomics, the study identifies four key volatile components linked to cordycepin, providing insights into developing functional rice porridge products. These findings are significant for advancing cordycepin mass production and offering dietary options for older individuals.

Complementary to Multiple Labels: A Correlation-Aware Correction Approach.

Complementary label learning (CLL) requires annotators to give irrelevant labels instead of relevant labels for instances. Currently, CLL has shown its promising performance on multi-class data by estimating a transition matrix. However, current multi-class CLL techniques cannot work well on multi-labeled data since they assume each instance is associated with one label while each multi-labeled instance is relevant to multiple labels. Here, we show theoretically how the estimated transition matrix in multi-class CLL could be distorted in multi-labeled cases as they ignore co-existing relevant labels. Moreover, theoretical findings reveal that calculating a transition matrix from label correlations in multi-labeled CLL (ML-CLL) needs multi-labeled data, while this is unavailable for ML-CLL. To solve this issue, we propose a two-step method to estimate the transition matrix from candidate labels. Specifically, we first estimate an initial transition matrix by decomposing the multi-label problem into a series of binary classification problems, then the initial transition matrix is corrected by label correlations to enforce the addition of relationships among labels. We further show that the proposal is classifier-consistent, and additionally introduce an MSE-based regularizer to alleviate the tendency of BCE loss overfitting to noises. Experimental results have demonstrated the effectiveness of the proposed method.

ZEB family is a prognostic biomarker and correlates with anoikis and immune infiltration in kidney renal clear cell carcinoma.

BACKGROUND: Zinc finger E-box binding homEeobox 1 (ZEB1) and ZEB2 are two anoikis-related transcription factors. The mRNA expressions of these two genes are significantly increased in kidney renal clear cell carcinoma (KIRC), which are associated with poor survival. Meanwhile, the mechanisms and clinical significance of ZEB1 and ZEB2 upregulation in KIRC remain unknown. METHODS: Through the Cancer Genome Atlas (TCGA) database and Gene Expression Omnibus (GEO) database, expression profiles, prognostic value and receiver operating characteristic curves (ROCs) of ZEB1 and ZEB2 were evaluated. The correlations of ZEB1 and ZEB2 with anoikis were further assessed in TCGA-KIRC database. Next, miRTarBase, miRDB, and TargetScan were used to predict microRNAs targeting ZEB1 and ZEB2, and TCGA-KIRC database was utilized to discern differences in microRNAs and establish the association between microRNAs and ZEBs. TCGA, TIMER, TISIDB, and TISCH were used to analyze tumor immune infiltration. RESULTS: It was found that ZEB1 and ZEB2 expression were related with histologic grade in KIRC patient. Kaplan-Meier survival analyses showed that KIRC patients with low ZEB1 or ZEB2 levels had a significantly lower survival rate. Meanwhile, ZEB1 and ZEB2 are closely related to anoikis and are regulated by microRNAs. We constructed a risk model using univariate Cox and LASSO regression analyses to identify two microRNAs (hsa-miR-130b-3p and hsa-miR-138-5p). Furthermore, ZEB1 and ZEB2 regulate immune cell invasion in KIRC tumor microenvironments. CONCLUSIONS: Anoikis, cytotoxic immune cell infiltration, and patient survival outcomes were correlated with ZEB1 and ZEB2 mRNA upregulation in KIRC. ZEB1 and ZEB2 are regulated by microRNAs.

Food-Borne Biotoxin Neutralization in Vivo by Nanobodies: Current Status and Prospects.

Food-borne biotoxins from microbes, plants, or animals contaminate unclean, spoiled, and rotten foods, posing significant health risks. Neutralizing such toxins is vital for human health, especially after food poisoning. Nanobodies (Nbs), a type of single-domain antibodies derived from the genetic cloning of a variable domain of heavy chain antibodies (VHHs) in camels, offer unique advantages in toxin neutralization. Their small size, high stability, and precise binding enable effective neutralization. The use of Nbs in neutralizing food-borne biotoxins offers numerous benefits, and their genetic malleability allows tailored optimization for diverse toxins. As nanotechnology continues to evolve and improve, Nbs are poised to become increasingly efficient and safer tools for toxin neutralization, playing a pivotal role in safeguarding human health and environmental safety. This review not only highlights the efficacy of these agents in neutralizing toxins but also proposes innovative solutions to address their current challenges. It lays a solid foundation for their further development in this crucial field and propels their commercial application, thereby contributing significantly to advancements in this domain.

SR9009 attenuates inflammation-related NPMSC pyroptosis and IVDD through NR1D1/NLRP3/IL-1ß pathway.

Intervertebral disc is a highly rhythmical tissue. As a key factor linking biorhythm and inflammatory response, the shielding effect of NR1D1 in the process of intervertebral disc degeneration remains unclear. Here, we first confirmed that NR1D1 in the nucleus pulposus tissue presents periodic rhythmic changes and decreases in expression with intervertebral disc degeneration. Second, when NR1D1 was activated by SR9009 in vitro, NLRP3 inflammasome assembly and IL-1ß production were inhibited, while ECM synthesis was increased. Finally, the vivo experiments further confirmed that the activation of NR1D1 can delay the process of disc degeneration to a certain extent. Mechanistically, we demonstrate that NR1D1 can bind to IL-1ß and NLRP3 promoters, and that the NR1D1/NLRP3/IL-1ß pathway is involved in this process. Our results demonstrate that the activation of NR1D1 can effectively reduce IL-1ß secretion, alleviate LPS-induced NPMSC pyroptosis, and protect ECM degeneration.

MiR-122 overexpression alleviates oxygen-glucose deprivation-induced neuronal injury by targeting sPLA2-IIA.

Background: Ischemic stroke (IS) is a neurological disease with significant disability and mortality. MicroRNAs were proven to be associated with cerebral ischemia. Previous studies have demonstrated miR-122 downregulation in both animal models of IS and the blood of IS patients. Nonetheless, the role and mechanism of miR-122-5p in IS remain unclear. Methods: We established primary human and mouse astrocytes, along with HT22 mouse hippocampal neuronal cells, through oxygen-glucose deprivation/reoxygenation (OGD/R) treatment. To assess the impact of miR-122, we employed CCK8 assays, flow cytometry, RT-qPCR, western blotting, and ELISA to evaluate cell viability, apoptosis, reactive oxygen species (ROS) generation, and cytokine expression. A dual-luciferase reporter gene assay was employed to investigate the interaction between miR-122 and sPLA2-IIA. Results: Overexpression of miR-122 resulted in decreased apoptosis, reduced cleaved caspase-3 expression, and increased cell viability in astrocytes and HT22 cells subjected to OGD/R. RT-qPCR and ELISA analyses demonstrated a decrease in mRNA and cytokine levels of interleukin (IL)-6 and tumor necrosis factor (TNF)-α in both astrocytes and HT22 cells following miR-122 overexpression. Moreover, miR-122 overexpression reversed OGD/R-induced ROS levels and 8-OHdG formation in astrocytes. Additionally, miR-122 overexpression decreased the mRNA and protein expression of inducible nitric oxide synthase (iNOS). Furthermore, we found that miR-122 attaches to the 3'-UTR of sPLA2-IIA, thereby downregulate its expression. Conclusion: Our study demonstrates that miR-122-mediated inhibition of sPLA2-IIA attenuates OGD/R-induced neuronal injury by suppressing apoptosis, alleviating post-ischemic inflammation, and reducing ROS production. Thus, the miR-122/sPLA2-IIA axis may represent a promising target for IS treatment.

Antibody Profiling of Pan-Cancer Viral Proteome Reveals Biomarkers for Nasopharyngeal Carcinoma Diagnosis and Prognosis.

Diagnosing, predicting disease outcome, and identifying effective treatment targets for virus-related cancers are lacking. Protein biomarkers have the potential to bridge the gap between prevention and treatment for these types of cancers. While it has been shown that certain antibodies against EBV proteins could be used to detect nasopharyngeal carcinoma (NPC), antibodies targeting are solely a tiny part of the about 80 proteins expressed by the EBV genome. Furthermore, it remains unclear what role other viruses play in NPC since many diseases are the result of multiple viral infections. For the first time, this study measured both IgA and IgG antibody responses against 646 viral proteins from 23 viruses in patients with NPC and control subjects using nucleic acid programmable protein arrays. Candidate seromarkers were then validated by ELISA using 1665 serum samples from three clinical cohorts. We demonstrated that the levels of five candidate seromarkers (EBV-BLLF3-IgA, EBV-BLRF2-IgA, EBV-BLRF2-IgG, EBV-BDLF1-IgA, EBV-BDLF1-IgG) in NPC patients were significantly elevated than controls. Additional examination revealed that NPC could be successfully diagnosed by combining the clinical biomarker EBNA1-IgA with the five anti-EBV antibodies. The sensitivity of the six-antibody signature at 95% specificity to diagnose NPC was comparable to the current clinically-approved biomarker combination, VCA-IgA, and EBNA1-IgA. However, the recombinant antigens of the five antibodies are easier to produce and standardize compared to the native viral VCA proteins. This suggests the potential replacement of the traditional VCA-IgA assay with the 5-antibodies combination to screen and diagnose NPC. Additionally, we investigated the prognostic significance of these seromarkers titers in NPC. We showed that NPC patients with elevated BLLF3-IgA and BDLF1-IgA titers in their serum exhibited significantly poorer disease-free survival, suggesting the potential of these two seromarkers as prognostic indicators of NPC. These findings will help develop serological tests to detect and treat NPC in the future.

Purification, identification and Cryo-EM structure of prostatic acid phosphatase in human semen.

Prostatic acid phosphatase (PAP) is a glycoprotein that plays a crucial role in the hydrolysis of phosphate ester present in prostatic exudates. It is a well-established indicator for prostate cancer due to its elevated serum levels in disease progression. Despite its abundance in semen, PAP's influence on male fertility has not been extensively studied. In our study, we report a significantly optimized method for purifying human endogenous PAP, achieving remarkably high efficiency and active protein recovery rate. This achievement allowed us to better analyze and understand the PAP protein. We determined the cryo-electron microscopic (Cryo-EM) structure of prostatic acid phosphatase in its physiological state for the first time. Our structural and gel filtration analysis confirmed the formation of a tight homodimer structure of human PAP. This functional homodimer displayed an elongated conformation in the cryo-EM structure compared to the previously reported crystal structure. Additionally, there was a notable 5-degree rotation in the angle between the α domain and α/ß domain of each monomer. Through structural analysis, we revealed three potential glycosylation sites: Asn94, Asn220, and Asn333. These sites contained varying numbers and forms of glycosyl units, suggesting sugar moieties influence PAP function. Furthermore, we found that the active sites of PAP, His44 and Asp290, are located between the two protein domains. Overall, our study not only provide an optimized approach for PAP purification, but also offer crucial insights into its structural characteristics. These findings lay the groundwork for further investigations into the physiological function and potential therapeutic applications of this important protein.

Student Loss: Towards the Probability Assumption in Inaccurate Supervision.

Noisy labels are often encountered in datasets, but learning with them is challenging. Although natural discrepancies between clean and mislabeled samples in a noisy category exist, most techniques in this field still gather them indiscriminately, which leads to their performances being partially robust. In this paper, we reveal both empirically and theoretically that the learning robustness can be improved by assuming deep features with the same labels follow a student distribution, resulting in a more intuitive method called student loss. By embedding the student distribution and exploiting the sharpness of its curve, our method is naturally data-selective and can offer extra strength to resist mislabeled samples. This ability makes clean samples aggregate tightly in the center, while mislabeled samples scatter, even if they share the same label. Additionally, we employ the metric learning strategy and develop a large-margin student (LT) loss for better capability. It should be noted that our approach is the first work that adopts the prior probability assumption in feature representation to decrease the contributions of mislabeled samples. This strategy can enhance various losses to join the student loss family, even if they have been robust losses. Experiments demonstrate that our approach is more effective in inaccurate supervision. Enhanced LT losses significantly outperform various state-of-the-art methods in most cases. Even huge improvements of over 50% can be obtained under some conditions.

Determining the Optimal Treatment for Idiopathic Clubfoot: A Network Meta-Analysis of Randomized Controlled Trials.

BACKGROUND: Clubfoot, or congenital talipes equinovarus deformity, is a common anomaly affecting the foot in infants. However, clinical equipoise remains between different interventions, especially those based on the Ponseti method. The aim of this study was to examine the clinical outcomes of the various interventions for treating idiopathic clubfoot. METHODS: Searches of the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase, Scopus, and CINAHL were conducted. Randomized controlled trials comparing different interventions, including the Ponseti method, accelerated Ponseti method, Ponseti method with botulinum toxin type A (Botox) injection, Ponseti method with early tibialis anterior tendon transfer (TATT), Kite method, and surgical treatment, were included. Network meta-analyses (NMAs) were conducted according to the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) reporting guidelines. The primary outcomes were the change in total Pirani score and maximal ankle dorsiflexion. Secondary outcomes were the number of casts, time in casts, and rates of tenotomy, total complications, relapse, adverse events, and additional required major surgery. RESULTS: Eleven randomized controlled trials involving 740 feet were included. According to the SUCRA (surface under the cumulative ranking curve)-based relative ranking, the Ponseti method was associated with the best outcomes in terms of Pirani score changes, maximal ankle dorsiflexion, number of casts, adverse events, and total complications, whereas the accelerated Ponseti method was associated with the best outcomes in terms of time in casts and tenotomy rate. Early TATT ranked best in terms of relapse rate. The Ponseti method with Botox injection was associated with the best outcomes in terms of the need for additional major surgery. CONCLUSIONS: The NMAs suggest that the Ponseti method is the optimal treatment overall, despite potential drawbacks such as longer time in casts and higher rates of tenotomy, relapse, and the need for additional surgery compared with other modified approaches. Therefore, clinicians should consider how treatments can be tailored individually. LEVEL OF EVIDENCE: Therapeutic Level I . See Instructions for Authors for a complete description of levels of evidence.

Single-Cell Transcriptome Atlas and Regulatory Dynamics in Developing Cotton Anthers.

Plant anthers are composed of different specialized cell types with distinct roles in plant reproduction. High temperature (HT) stress causes male sterility, resulting in crop yield reduction. However, the spatial expression atlas and regulatory dynamics during anther development and in response to HT remain largely unknown. Here, the first single-cell transcriptome atlas and chromatin accessibility survey in cotton anther are established, depicting the specific expression and epigenetic landscape of each type of cell in anthers. The reconstruction of meiotic cells, tapetal cells, and middle layer cell developmental trajectories not only identifies novel expressed genes, but also elucidates the precise degradation period of middle layer and reveals a rapid function transition of tapetal cells during the tetrad stage. By applying HT, heterogeneity in HT response is shown among cells of anthers, with tapetal cells responsible for pollen wall synthesis are most sensitive to HT. Specifically, HT shuts down the chromatin accessibility of genes specifically expressed in the tapetal cells responsible for pollen wall synthesis, such as QUARTET 3 (QRT3) and CYTOCHROME P450 703A2 (CYP703A2), resulting in a silent expression of these genes, ultimately leading to abnormal pollen wall and male sterility. Collectively, this study provides substantial information on anthers and provides clues for heat-tolerant crop creation.

Biochemical and Transcriptome Analysis Reveals Pigment Biosynthesis Influenced Chlorina Leaf Formation in Anoectochilus roxburghii (Wall.) Lindl.

Anoectochilus roxburghii (Wall.) Lindl is a perennial herb of the Orchidaceae family; a yellow-green mutant and a yellow mutant were obtained from the wild type, thereby providing good material for the study of leaf color variation. Pigment content analysis revealed that chlorophyll, carotenoids, and anthocyanin were lower in the yellow-green and yellow mutants than in the wild type. Transcriptome analysis of the yellow mutant and wild type revealed that 78,712 unigenes were obtained, and 599 differentially expressed genes (120 upregulated and 479 downregulated) were identified. Using the Kyoto Encyclopedia of Genes and Genomes pathway analysis, candidate genes involved in the anthocyanin biosynthetic pathway (five unigenes) and the chlorophyll metabolic pathway (two unigenes) were identified. Meanwhile, the low expression of the chlorophyll and anthocyanin biosynthetic genes resulted in the absence of chlorophylls and anthocyanins in the yellow mutant. This study provides a basis for similar research in other closely related species.

Regulatory controls of duplicated gene expression during fiber development in allotetraploid cotton.

Polyploidy complicates transcriptional regulation and increases phenotypic diversity in organisms. The dynamics of genetic regulation of gene expression between coresident subgenomes in polyploids remains to be understood. Here we document the genetic regulation of fiber development in allotetraploid cotton Gossypium hirsutum by sequencing 376 genomes and 2,215 time-series transcriptomes. We characterize 1,258 genes comprising 36 genetic modules that control staged fiber development and uncover genetic components governing their partitioned expression relative to subgenomic duplicated genes (homoeologs). Only about 30% of fiber quality-related homoeologs show phenotypically favorable allele aggregation in cultivars, highlighting the potential for subgenome additivity in fiber improvement. We envision a genome-enabled breeding strategy, with particular attention to 48 favorable alleles related to fiber phenotypes that have been subjected to purifying selection during domestication. Our work delineates the dynamics of gene regulation during fiber development and highlights the potential of subgenomic coordination underpinning phenotypes in polyploid plants.