Downstream Applications Research Articles

Cloning, characterization of β-glucosidase from Furfurilactobacillus rossiae in bioconversion and its efficacy.

Minor ginsenosides produced by β-glucosidase are interesting biologically and pharmacologically. In this study, new ginsenoside-hydrolyzing glycosidase from Furfurilactobacillus rossiae DCYL3 was cloned and expressed in Escherichia coli strain BL21. The enzyme converted Rb1 and Gyp XVII into Rd and compound K following the pathways: Rb1→Rd and Gyp XVII→F2→CK, respectively at optimal condition: 40°C, 15min, and pH 6.0. Furthermore, we examined the cytotoxicity, NO production, ROS generation, and gene expression of Gynostemma extract (GE) and bioconverted Gynostemma extract (BGE) in vitro against A549 cell lines for human lung cancer and macrophage RAW 264.7 cells for antiinflammation, respectively. As a result, BGE demonstrated significantly greater toxicity than GE against lung cancer at a dose of 500µg/mL but in normal cells showed lower toxicity. Then, we indicated an enhanced generation of ROS, which may be boosting cancer cell toxicity. By blocking the intrinsic way, BGE increased p53, Bax, Caspase 3, 9, and while Bcl2 is decreased. At 500µg/mL, the BGE sample was less toxic in normal cells and decreased the LPS-treated NO and ROS level to reduce inflammation. In addition, BGE inhibited the expression of pro-inflammatory genes COX-2, iNOS, IL-6, and IL-8 in RAW 264.7 cells than the sample of GE. In conclusion, FrBGL3 has considerable downstream applications for high-yield, low-cost, effective manufacture of minor ginsenosides. Moreover, the study's findings imply that BGE would be potential materials for anti-cancer and anti-inflammatory agent after consideration of future studies.

Human sensory-like neuron cultivation-An optimized protocol.

Reprogramming of human-induced pluripotent stem cells (iPSCs) and their differentiation into specific cell types, such as induced sensory-like neurons (iSNs), are critical for disease modeling and drug testing. However, the variability of cell populations challenges reliability and reproducibility. While various protocols for iSN differentiation exist, the development of non-iSN cells in these cultures remains an issue. Therefore, standardization of protocols is essential. This study aimed to improve iSN culture conditions by reducing the number of non-iSN cells while preserving the survival and quality of iSNs. iSNs were differentiated from a healthy control iPSC line using an established protocol. Interventions for protocol optimization included floxuridine (FdU) or 1-β-D-arabinofuranosyl-cytosine hydrochloride (AraC) treatment, magnetic-activated cell sorting (MACS), early cell passaging, and replating. Cell viability and iSN-to-total-cell-count ratio were assessed using a luminescent assay and immunocytochemistry, respectively. Passaging of cells during differentiation did not increase the iSN-to-total-cell-count ratio, and MACS of immature iSNs led to neuronal blebbing and reduced the iSN-to-total-cell-count ratio. Treatment with high concentrations and prolonged incubation of FdU or AraC resulted in excessive cell death. However, treatment with 10 μM FdU for 24 h post-differentiation showed the most selective targeting of non-iSN cells, leading to an increase in the iSN-to-total-cell count ratio without compromising the viability or functionality of the iSN population. Replating of iSNs shortly after seeding also helped to reduce non-iSN cells. In direct comparison with other methods, treatment with 10 μM FdU for 24 h after differentiation shows promise for improving iSN culture purity, which could benefit downstream applications in disease modeling and drug discovery. However, further investigations involving multiple iPSC lines and optimization of protocol parameters are warranted to fully exploit the potential of this method and enhance its reproducibility and applicability. Overall, this study provides valuable insights into optimizing culture conditions for iSN differentiation and highlights the importance of standardized protocols in iPSC-based research.

B-185 A Comparison Between Two Oral Sample Collection Devices on Preserving High-Quality Genomic DNA

Abstract Background High-quality genomic testing significantly enhances the knowledge on gene-disease association, and greatly benefits future biomedical research and precision medicine. However, widely used saliva sample collection presents challenges with high-quality and quantity of human DNA and high bacterial DNA ratio. In this study, we validated the performance of two commercially available oral collection devices for their ability to preserve high-quality genomic DNA long-term at room temperature. Methods Fifty-one buccal samples in total from twelve donors were collected using two kinds of commercially available oral sample collection devices (iSWAB®-DNA (Device 1) and OCD-100 (Device 2)). All of the collected samples were stored at ambient temperature for up to 1 year. Collected samples were extracted with QIAamp Blood Mini Kit (Qiagen) at baseline, 30 days, 60 days, and 1 year. All DNA yields were measured using a Nanodrop One Microvolume UV-Vis Spectrophotometer; TapeStation System was used for determining the integrity of DNA. The bacterial DNA ratio was estimated by quantitative PCR with primers targeting the 16S rRNA gene known to be present in prokaryotes but not eukaryotes. Results Device 1 had higher yield (Device 1: 18.59 ± 8.98; Device 2: 6.44 ± 4.62), lower bacterial DNA ratio (Device 1: 2.06 ± 1.93; Device 2: 3.79 ± 4.58), and higher overall DNA integrity (Device 1: 6.1 ± 0.31; Device 2: 5.42 ± 0.96) compared to Device 2. Device 1 demonstrated longer post-collection sample stability at ambient temperature compared with Device 2 regarding DNA yield, bacterial DNA ratio and DNA integrity. No significant changes in DNA yield, integrity, and bacterial DNA ratio were observed after one year of storage in Device 1. No significant changes in DNA yield, integrity, and bacterial DNA ratio were observed after 60 days of storage in Device 2. However, Device 2 showed a significant increase in DNA yield, bacterial DNA ratio, and DNA integrity comparing baseline to 1 year. Conclusions Device 1 preserves higher quantity and quality of DNA from oral samples at ambient temperature for up to 1 year. These advantages significantly enhance efficiency and cost-effectiveness in downstream applications such as genotyping and sequencing, making it a valuable tool in genomics research with a commendable success rate.

B-184 Purifying High-Quality Genomic DNA From Frozen Blood Samples Stored up to 1.5 Years at -20°C

Abstract Background Maintaining DNA integrity in blood samples, from transport to storage to processing, is crucial to the success of downstream applications, especially regarding diagnostic applications. Additionally, due to the invasive nature of collecting blood samples, minimizing resampling to avoid excess patient pain and cost of collection and transportation are important to consider. Having the option to store and maintain portions of blood samples in the freezer enables future retesting without recollection. Since -20°C freezers generally are more affordable, require less energy, and have smaller footprints than -80°C freezers, demonstrating frozen blood stability at -20°C temperatures could decrease sampling burdens for labs with facility constraints. Methods Whole blood was collected into 9mL vacuum tubes (HemaSure-OMXP) from healthy donors. Aliquots of whole blood and buffy coat samples were stored at -20°C and thawed for genomic DNA extractions (QIAamp DNA Blood Mini Kit, Qiagen) at a range of different timepoints; frozen whole blood samples were processed at 0, 49, 141, and 148 days and frozen buffy coat samples were processed at 7, 27, 36, 185, 246, 373, 394, 584, and 604 days. Purified gDNA was quantified with Qubit (Broad Range dsDNA Assay, ThermoFisher) and qualified with Nanodrop One Microvolume UV-Vis Spectrophotometer (ThermoFisher) and TapeStation (Genomic DNA ScreenTape, Agilent Technologies). Results No significant decrease was observed for DNA yield and quality between fresh and frozen whole blood samples. Average DNA yields from 200 μL whole blood were 4.5 ± 2.2 μg and 5.3 ± 1.2 μg for fresh and frozen blood respectively. Average DNA integrity (DIN) was 6.9 ± 0.4 and 7.1 ± 0.3 for fresh and frozen blood respectively. Donor differences in blood samples can help explain the variance in DNA yields reported. No significant decrease was observed for DNA yield and quality in buffy coat samples frozen for longer at -20°C. In total, 84 whole blood and 25 buffy coat samples were processed in this study. Conclusions Although buffy coat samples tended to have higher yields than whole blood, due to the higher amount of white blood cells present, both sample types provided sufficient amounts of gDNA for downstream applications. Freezing blood samples for future testing can be effective at reducing the need for resampling without sacrificing the nucleic acid yield or quality of fresh samples. Additionally, as technological advancements are made for equipment and assays, frozen blood samples can be retested and compared to previous results for improved diagnostic decisions, done without requiring recollection from labs and patients.

B-022 A New DNA Size-Selection Technology:Methods for Isolating, Capturing, and Recovery of Target Macromolecules by Gel Electrophoresis

Abstract Background The ability to recover DNA fragments in high quantity and purity from agarose gel is an important need for researchers in molecular biology laboratories for downstream applications such as testing and sequencing. There are multiple established methods and corresponding commercial products available at researchers’ disposal. Despite the availability of many methods, researchers often face a trade-off decision between fast and simple methods with low yield and poor DNA quality, and time-consuming and expensive methods with higher yields and better purity. Methods Gene Stix (ElectroSep) is a new system for recovery of DNA/RNA from an electrophoresis gel. Using the power of electrophoresis to separate nucleic acids based on size (a well-known technique), the Gene Stix, for the first time, allows one to recover nucleic acids directly from the gel. Gene Stix consisted of a cationic membrane which is inserted (in front of the nucleic acid of interest) into the electrophoresis gel after the nucleic acids are resolved from other contaminants in a separate step. After the Gene Stix is inserted in front of the DNA, the gel is re-run for a short period of time, binding the DNA to the Gene Stix. The Gene Stix is removed from the gel (carrying the DNA) and the DNA is recovered by elution using a special buffer. Results Electro-Sep technology allows DNA recovery around 85% and may lead up to 35% process efficiency for purification. It maximizes the process efficiency by not requiring damaging UV sensitive dyes in samples or cutting gel bands and allowing multiple Gene Stix used with one sample. Unique serial numbering system provides unambiguous sample identification. Conclusions The yield from different methods varies, however, it is typically at around 60-70%. Many of the commercially available DNA purification kits require around 25-30 minutes for purification. Moreover, the process with available methods could be laborious requiring cutting the gel bands, using UV dyes in sample, or simply requiring specialized gel electrophoresis equipment. Electro-Sep technology (and Gene Stix product using the technology), designed to recover size-selected DNA from agarose gel, provides a viable option that is rapid, easy-to-use yet resulting in high DNA recovery.

DNARS: A safe, environmentally friendly and high-quality DNA extraction method suitable for various biological samples

The functional and interactional relationships between nucleic acids and proteins form the foundational framework for molecular biology studies. In this context, obtaining high-quality nucleic acids is a crucial step for successful downstream applications. This study aimed to develop a cost-effective, eco-friendly, and versatile DNA extraction method that yields high-quality DNA from diverse biological samples. We utilized three types of borosilicate-based recycled laboratory glass as a silica source, combined with sodium iodide as a chaotropic agent, creating an efficient system for cell lysis and DNA extraction. Quality control was performed by assessing the concentration and purity of the extracted DNA using spectrophotometry, and the results were compared to those obtained with a commercial kit. DNA integrity was evaluated via agarose gel electrophoresis. To verify the suitability of the extracted DNA for downstream applications, we conducted 16S rRNA and ITS PCR amplifications. Our findings demonstrated that our DNA extraction method produced significantly higher yields, better purity, and greater integrity compared to the commercial kit. Moreover, the extracted DNA was readily applicable in PCR-based procedures, confirming the method's effectiveness for molecular biology applications.

Human monoclonal antibody cloning and expression with overlap extension PCR and short DNA fragments

Monoclonal antibodies are powerful therapeutic, diagnostic, and research tools. Methods utilized to generate monoclonal antibodies are evolving rapidly. We created a transfectable linear antibody expression cassette from a 2-h high-fidelity overlapping PCR reaction from synthesized DNA fragments. We coupled heavy and light chains into a single linear sequence with a promoter, self-cleaving peptide, and poly(A) signal to increase the flexibility of swapping variable regions from any sequence available in silico. Transfection of the linear cassette tended to generate similar levels to the two-plasmid system and generated an average of 47 μg (14–98 μg) after 5 days in 2 mL cultures with 15 unique antibody sequences. The levels of antibodies produced were sufficient for most downstream applications in less than a week. The method presented here reduces the time, cost, and complexity of cloning steps.

GEDI: An R Package for Integration of Transcriptomic Data from Multiple Platforms for Bioinformatics Applications.

Transcriptomic data is often expensive and difficult to generate in large cohorts relative to genomic data; therefore, it is often important to integrate multiple transcriptomic datasets from both microarray- and next generation sequencing (NGS)-based transcriptomic data across similar experiments or clinical trials to improve analytical power and discovery of novel transcripts and genes. However, transcriptomic data integration presents a few challenges including reannotation and batch effect removal. We developed the Gene Expression Data Integration (GEDI) R package to enable transcriptomic data integration by combining existing R packages. With just four functions, the GEDI R package makes constructing a transcriptomic data integration pipeline straightforward. Together, the functions overcome the complications in transcriptomic data integration by automatically reannotating the data and removing the batch effect. The removal of the batch effect is verified with principal component analysis and the data integration is verified using a logistic regression model with forward stepwise feature selection. To demonstrate the functionalities of the GEDI package, we integrated five bovine endometrial transcriptomic datasets from the NCBI Gene Expression Omnibus. These transcriptomic datasets were from multiple high-throughput platforms, namely, array-based Affymetrix and Agilent platforms, and NGS-based Illumina paired-end RNA-seq platform. Furthermore, we compared the GEDI package to existing tools and found that GEDI is the only tool that provides a full transcriptomic data integration pipeline including verification of both batch effect removal and data integration for downstream genomic and bioinformatics applications. © 2024 The Author(s). Current Protocols published by Wiley Periodicals LLC. Basic Protocol 1: ReadGE, a function to import gene expression datasets Basic Protocol 2: GEDI, a function to reannotate and merge gene expression datasets Basic Protocol 3: BatchCorrection, a function to remove batch effects from gene expression data Basic Protocol 4: VerifyGEDI, a function to confirm successful integration of gene expression data.

Multi-level discriminator based contrastive learning for multiplex networks

Graph embedding is a technique for obtaining low-dimensional representations of nodes across diverse networks, which may then be used for various downstream tasks and applications. When it applies to heterogeneous networks, it is hard to handle heterogeneous networks because they usually contain different types of nodes and edges with more semantic and structural information. Recently, contrastive learning has developed as the preferred strategy for dealing with unsupervised heterogeneous graph embedding to reduce the cost of human label annotation. However, most multi-view contrastive learning approaches calculate the model’s loss only based on the mutual dependence between the node representation and graph representation. These approaches ignore that both node attributes and node clustering contain discriminative content. To solve this issue, we propose a model called Multi-Level Discriminator-based Contrastive Learning for Multiplex Networks (MLDCL). This model adopts a multi-level multi-discriminator-based approach that can simultaneously learn the global-level structural information, node-level attribute information, and local-level clustering information. Moreover, an augmentation strategy in the contrast learning process from the spectral domain is proposed to improve the representation and discriminative ability of MLDCL. Numerous tests with node clustering and classification tasks on widely used datasets demonstrate the efficacy of the proposed approach.

Interactions between antibiotic removal, water matrix characteristics and layered double hydroxide sorbent material

Sorption by layered double hydroxides (LDH) is gaining substantial interest for remediating emerging contaminants, including pharmaceuticals from wastewaters. Findings from a sorbent material performing successfully in lab-based studies using non-environmental (laboratory-sourced) water cannot be assumed to translate to equal performance under environmental downstream applications. However, studies evaluating sorbent material performance for removal of pollutants and understanding material interactions with environmental waters are limited. This study evaluates the removal of the antibiotic amoxicillin (AMX) using a Mg2Al–NO3-LDH sorbent material from laboratory-grade water and wastewater effluent (WWE). AMX is successfully removed (94.53 ± 4.30 % within 24 h) in laboratory-grade water (under batch sorption conditions: 100 μg/L AMX, 0.2 g/L LDH, 20 °C). The comparison of LDH removal performance in laboratory grade and WWE shows a decreased maximum removal of AMX in WWE (13.39 ± 5.53 %). A lower final AMX concentration is observed in the WWE without the presence of LDH, compared to the ‘removal’ experiments in WWE with the presence of LDH, indicating a contribution of non-sorption removal pathways of AMX. This is proposed to be due to the difference in metal concentrations in the WWE with and without LDH present. The presence of LDH is found to decrease concentrations of metal pollutants in WWE, such as Zn concentration decreasing by 85 % over 24 h, changing water characteristics. Overall, this paper reports that an LDH performs differently in laboratory-sourced water and a wastewater effluent. This provides evidence that sorbent material performance needs to be evaluated in complex water matrices to ensure that it is representative of how a sorbent material will perform in an environmental application, which is the end goal of developing such technologies. Finally, good practice recommendations are provided for future lab-scale sorption experiments evaluating the performance of any new sorbent materials for water treatment applications.

SWQC: Efficient sequencing data quality control on the next-generation sunway platform

Sequencing data quality control can significantly prevent low-quality data from impacting downstream applications in bioinformatics. The enormous growth of biological sequencing data in recent years introduces new challenges to the efficiency of quality control processes and motivates the need for fast implementations on modern compute systems. The powerful next-generation heterogeneous Sunway platform holds significant potential for addressing this challenge. However, there are currently no dedicated quality control applications that can fully utilize its computational power. To bridge this gap, we introduce SWQC, a novel quality control application specifically designed for the Sunway platform. We present an efficient distributed FASTQ I/O framework for Sunway-based workstations and supercomputers to take advantage of fast SSDs and the parallel file system. In order to support both process-level and thread-level (CPE-level) parallelism to leverage the computational power, we refactor and optimize all standard quality control modules for the heterogeneous Sunway architecture. When using a single node, SWQC achieves speedups between 2 and 40 over highly optimized quality control applications executed on a high-end 48-core AMD server. Additionally, when using 16 nodes, SWQC achieves parallel efficiencies of 70% (for reading and writing a single file) and 95% (for reading one file and writing split files) compared to a single node. Overall, SWQC is able to perform quality control operations for a 140GB FASTQ file within only 70 s using a single Sunway node. It is publicly available at https://github.com/RabbitBio/SWQC.

Importance-Induced Customer Segmentation Using Explainable Machine Learning

Abstract Customer segmentation plays a critical role in enhancing a company's product penetration rate in the market. It enables numerous downstream applications such as customer-oriented product development and trend analysis. Previous approaches to customer segmentation have relied either on survey-based methods or data-driven approaches. However, these methods face challenges such as high human labor requirements or the generation of noisy segments. To address these challenges, this paper proposes a new methodology based on data-driven network construction and an importance-enhanced framework. The framework incorporates two techniques: (1) the utilization of a neural network model to compute feature importance values and (2) the proposal of a novel network connection rule. This framework addresses the limitation of the previous approach, sentimentpolarity- based networking, by connecting customers based on feature importance. We further validated the effectiveness of the framework using three real-world datasets and demonstrated the proposed method outperformed the previous approach.

Building an Ethical and Trustworthy Biomedical AI Ecosystem for the Translational and Clinical Integration of Foundation Models.

Foundation Models (FMs) are gaining increasing attention in the biomedical artificial intelligence (AI) ecosystem due to their ability to represent and contextualize multimodal biomedical data. These capabilities make FMs a valuable tool for a variety of tasks, including biomedical reasoning, hypothesis generation, and interpreting complex imaging data. In this review paper, we address the unique challenges associated with establishing an ethical and trustworthy biomedical AI ecosystem, with a particular focus on the development of FMs and their downstream applications. We explore strategies that can be implemented throughout the biomedical AI pipeline to effectively tackle these challenges, ensuring that these FMs are translated responsibly into clinical and translational settings. Additionally, we emphasize the importance of key stewardship and co-design principles that not only ensure robust regulation but also guarantee that the interests of all stakeholders-especially those involved in or affected by these clinical and translational applications-are adequately represented. We aim to empower the biomedical AI community to harness these models responsibly and effectively. As we navigate this exciting frontier, our collective commitment to ethical stewardship, co-design, and responsible translation will be instrumental in ensuring that the evolution of FMs truly enhances patient care and medical decision-making, ultimately leading to a more equitable and trustworthy biomedical AI ecosystem.

Toward subtask-decomposition-based learning and benchmarking for predicting genetic perturbation outcomes and beyond.

Deciphering cellular responses to genetic perturbations is fundamental for a wide array of biomedical applications. However, there are three main challenges: predicting single-genetic-perturbation outcomes, predicting multiple-genetic-perturbation outcomes and predicting genetic outcomes across cell lines. Here we introduce Subtask Decomposition Modeling for Genetic Perturbation Prediction (STAMP), a flexible artificial intelligence strategy for genetic perturbation outcome prediction and downstream applications. STAMP formulates genetic perturbation prediction as a subtask decomposition problem by resolving three progressive subtasks in a problem decomposition manner, that is, identifying postperturbation differentially expressed genes, determining the expression change directions of differentially expressed genes and finally estimating the magnitudes of gene expression changes. STAMP exhibits a substantial improvement over the existing approaches on three subtasks and beyond, including the ability to identify key regulatory genes and pathways on small samples and to reveal precise genetic interactions of diverse types.

Knowledge graph of agricultural engineering technology based on large language model

Agriculture is an industry that has evolved alongside human evolution and has faithfully fulfilled its core mission of food supply. With the reduction of rural labor, the progress of artificial intelligence and the development of Internet of Things technology, it is hoped that the efficiency and productivity of the agricultural industry can be improved. Recently, with the development of information and intelligent technology, agricultural production and management have been significantly enhanced. However, there is still a considerable challenge in effectively integrating the vast amount of fragmented information for downstream applications. An agricultural knowledge graph (AGKG) will serve as the foundation for achieving these goals. Knowledge graphs can be general or domain-specific, and are the basis for many applications, such as search engines, online question-and-answer services, and knowledge inference. Therefore, there are many knowledge graphs, including Wikidata and DBpedia, for accessing structured knowledge. Although some general knowledge graphs contain some entities and relationships related to agriculture, there are no domain-specific knowledge graphs specifically for agricultural applications. Therefore, this paper proposes an agricultural knowledge graph (AGKG) for automatically integrating large amounts of agricultural data from the Internet. By applying natural language processing and deep learning technologies, AGKG can automatically identify agricultural entities from unstructured text and connect them to form a knowledge graph. In addition, we have described the typical scenarios of our AGKG and validated it through real-world applications such as agricultural entity retrieval and agricultural question-answering.

A survey on pragmatic processing techniques

Pragmatics, situated in the domains of linguistics and computational linguistics, explores the influence of context on language interpretation, extending beyond the literal meaning of expressions. It constitutes a fundamental element for natural language understanding in machine intelligence. With the advancement of large language models, the research focus in natural language processing has predominantly shifted toward high-level task processing, inadvertently downplaying the importance of foundational pragmatic processing tasks. Nevertheless, pragmatics serves as a crucial medium for unraveling human language cognition. The exploration of pragmatic processing stands as a pivotal facet in realizing linguistic intelligence. This survey encompasses important pragmatic processing techniques for subjective and emotive tasks, such as personality recognition, sarcasm detection, metaphor understanding, aspect extraction, and sentiment polarity detection. It spans theoretical research, the forefront of pragmatic processing techniques, and downstream applications, aiming to highlight the significance of these low-level tasks in advancing natural language understanding and linguistic intelligence.

Representation Learning for Sequential Volumetric Design Tasks

Abstract Volumetric design, also called massing design, is the first and critical step in professional building design which is sequential in nature. As the volumetric design process is complex, the underlying sequential design process en- codes valuable information for designers. Many efforts have been made to automatically generate reasonable volumetric designs, but the quality of the generated design solutions varies, and evaluating a design solution requires either a prohibitively comprehensive set of metrics or expensive human expertise. While previous approaches focused on learning only the final design instead of sequential design tasks, we propose to encode the design knowledge from a collection of expert or high-performing design sequences and extract useful representations using transformer-based models. Later we propose to utilize the learned representations for crucial downstream applications such as design preference evaluation and procedural design generation. We develop the preference model by estimating the density of the learned representations whereas we train an autoregressive transformer model for sequential design generation. We demonstrate our ideas by leveraging a novel dataset of thousands of sequential volumetric designs. Our preference model can compare two arbitrarily given design sequences and is almost 90% accurate in evaluation against random design sequences. Our autoregressive model is also capable of autocompleting a volumetric design sequence from a partial design sequence.

ECAD-Net: Editable Parametric CAD Models Reconstruction from Dumb B-Rep Models Using Deep Neural Networks

This paper introduces a novel framework capable of reconstructing editable parametric CAD models from dumb B-Rep models. First, each B-Rep model is represented with a network-friendly formalism based on UV-graph, which is then used as input of eCAD-Net, the new deep neural network-based algorithm that predicts feature-based CAD modeling sequences from the graph. Then, the sequences are scaled and fine-tuned using a feature matching algorithm that retrieves the exact parameter values from the input dumb CAD model. The output sequences are then converted in a series of CAD modeling operations to create an editable parametric CAD model in any CAD modeler. A cleaned dataset is used to learn and validate the proposed approach, and is provided with the article. The experimental results show that our approach outperforms existing methods on such reconstruction tasks, and it outputs editable parametric CAD models compatible with existing CAD modelers and ready for use in downstream engineering applications.

Estimation of the Uncertainty in Daytime Cirrus Cloud Radiative Forcing and Heating Rates Due to Ice Crystal Optics

Abstract The uncertainties in absolute daytime top-of-the-atmosphere (TOA) net cirrus cloud radiative forcing (CRF) and radiative heating rates are estimated at five Micro-Pulse Lidar Network (MPLNET) sites spanning the tropics to high-latitudes. One year of semi-transparent cirrus cloud (optical depth < 3.0 and cloud top temperature < −37 °C) measurements are subject to spectrally-consistent optical properties for nine different ice crystal habits, thus providing a range of possible forcing values. The annual average absolute daytime TOA net CRF is positive at Barbados, Kanpur, and, Singapore (0.59–0.67, 0.61–0.65, and 1.94–2.09 W∙m−2, respectively), negative at Fairbanks (−0.67 to −0.28 W∙m−2), and can regularly become positive or negative at Goddard Space Flight Center (GSFC) (−0.06 to 0.32 W∙m−2). The TOA CRF depends on ice crystal shape; in particular, plates lead to relatively large absolute values that decreases for bullet rosettes and columns. Uncertainties in daytime cirrus cloud radiative properties are estimated as the standard deviation of all possible outcomes when considering the different particle habits individually. Annually, the average uncertainty of the absolute daytime TOA net CRF ranges from 0.50–1.80 W∙m−2. In-cloud daytime net radiative heating rates are positive, on average, at all five sites (0.25–3.84 K/day) and have an estimated uncertainty of less than 0.30 K/day. The uncertainties in cirrus radiative forcing and heating that are characterized by assumptions regarding the ice crystal optical properties must be considered in downstream applications, including satellite retrievals and numerical weather prediction.

DHICA: a deep transformer-based model enables accurate histone imputation from chromatin accessibility

Abstract Histone modifications (HMs) are pivotal in various biological processes, including transcription, replication, and DNA repair, significantly impacting chromatin structure. These modifications underpin the molecular mechanisms of cell-type-specific gene expression and complex diseases. However, annotating HMs across different cell types solely using experimental approaches is impractical due to cost and time constraints. Herein, we present dHICA (deep histone imputation using chromatin accessibility), a novel deep learning framework that integrates DNA sequences and chromatin accessibility data to predict multiple HM tracks. Employing the transformer architecture alongside dilated convolutions, dHICA boasts an extensive receptive field and captures more cell-type-specific information. dHICA outperforms state-of-the-art baselines and achieves superior performance in cell-type-specific loci and gene elements, aligning with biological expectations. Furthermore, dHICA’s imputations hold significant potential for downstream applications, including chromatin state segmentation and elucidating the functional implications of SNPs (Single Nucleotide Polymorphisms). In conclusion, dHICA serves as a valuable tool for advancing the understanding of chromatin dynamics, offering enhanced predictive capabilities and interpretability.