Pipeline Approach Research Articles

Pipelined and Wave Pipelined Approach Based Comparative Analysis for 16x16 Vedic Multiplier

Objectives: This work objective is to construct an FPGA-based 16x16 Vedic multiplier and assess the performance of the multiplier using three distinct architectures: pipeline, wave pipeline, and modified wave pipeline in terms of delay and clock skew. Methods: The 16 × 16 Vedic multiplier was constructed and designed through four numbers of an 8x8 Vedic multiplier. For the 16x16 Vedic multiplier, the 3-stage pipeline and wave pipeline techniques are applied, and the delay performances of the Vedic multiplier are compared. Delay optimization: In the wave pipeline Vedic multiplier architecture, the delay is decreased by inserting the known delay on the longest path delay between the multiplier and adder. Clock skew optimization: The clock skew issue of the wave pipeline Vedic multiplier architecture is minimized by adjusting the setup time violation of the clock signal that is connected to the input and output registers. Findings: The delay performance of the Vedic multiplier was evaluated by the synthesis tools Xilinx 12.1, Xilinx ISE 14.2, and Altera, and based on the synthesis report, the Xilinx synthesis tool offers 73.71% delay performance for the pipeline approach and 53.39% for the wave pipeline approach compared to the Altera tool. Further delay is reduced by the proposed modified wave pipeline approach, which saves 2.122 ns of delay compared to the wave pipeline architecture. The clock skew performance was analyzed using the Time Quest timing analyzer tool, and it was minimized to 0.035 from 0.048 compared to the wave pipeline approach. Novelty: In this work, the modified wave pipeline approach has been applied to the existing Vedic multiplier architecture, and it offers less delay as well as less clock skew compared to the existing method. Hence, the performance of the Vedic multiplier with a modified wave pipelined approach was evaluated through a 3-tap FIR filter by applying a vibroarthrography signal. Keywords: Pipeline, Wave Pipeline, Vedic Multiplier, Clock skew, Set up violation, Altera quartex- II Time quest timing analyzer tool

Fast labeling pipeline approach for a huge aerial sensed dataset

Fast labeling pipeline approach for a huge aerial sensed dataset

Revisiting clustering for efficient unsupervised dialogue structure induction

In the development of a task-oriented dialogue system, defining the dialogue structure is a time-consuming task. Hence, several works have looked into automatically inferring it from data, e.g., actual conversations between a customer and a support agent. To recover such dialogue structure, recent methods based on discrete variational models learn to jointly encode and cluster utterances in dialogue states, but (i) represent utterances by only considering preceding dialogue context, and (ii) are slow to train since they are optimized with a compute-expensive decoding objective. We revisit and improve upon an existing efficient pipeline approach, commonly adopted as a baseline, that first encodes utterances and then clusters them with k-means to induce the dialogue structure. However, the existing approach represents utterances as bag-of-words or skip-thought vectors, which have been shown to perform poorly in semantic similarity tasks, and without considering dialogue context. We therefore first investigate the use of more powerful transformer-based encoders for encoding utterances. Next, we propose ellodar, a method for learning representations that capture both preceding and subsequent dialogue context, inspired by word-to-vec training strategies. ellodar is efficient since representations are learned directly in the encoding space by finetuning just a single linear layer on top of a frozen sentence encoder with a vector-to-vector regression training objective. Extensive experiments on representative datasets for dialogue structure induction (SimDial, Schema Guided Dialogues, DSTC2, and CamRest676) demonstrate that in terms of effectiveness to induce the correct dialogue structure, (i) clustering utterances represented by transformed-based encoders improves recent joint models by 13%–32% on standard cluster metrics, and (ii) clustering ellodar’s representations yields additional improvements ranging from +20% to +26%, with speedups of ×\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ imes $$\\end{document} 10\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extbf{10}$$\\end{document}–104\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ extbf{10}^{\ extbf{4}}$$\\end{document} compared to the recent joint models.

Enhance Speed Low Area FPGA Design Using S-Box GF and Pipeline Approach on Logic for AES

Enhance Speed Low Area FPGA Design Using S-Box GF and Pipeline Approach on Logic for AES

Improved Regression Analysis with Ensemble Pipeline Approach for Applications across Multiple Domains

In this research, we introduce two new machine learning regression methods: the Ensemble Average and the Pipelined Model. These methods aim to enhance traditional regression analysis for predictive tasks and have undergone thorough evaluation across three datasets, Kaggle House Price, Boston House Price, and California Housing, using various performance metrics. The results consistently show that our models outperform existing methods in terms of accuracy and reliability across all three datasets. The Pipelined Model, in particular, is notable for its ability to combine predictions from multiple models, leading to higher accuracy and impressive scalability. This scalability allows for their application in diverse fields like technology, finance, and healthcare. Furthermore, these models can be adapted for real-time and streaming data analysis, making them valuable for applications such as fraud detection, stock market prediction, and IoT sensor data analysis. Enhancements to the models also make them suitable for big data applications, ensuring their relevance for large datasets and distributed computing environments. It is important to acknowledge some limitations of our models, including potential data biases, specific assumptions, increased complexity, and challenges related to interpretability when using them in practical scenarios. Nevertheless, these innovations advance predictive modeling, and our comprehensive evaluation underscores their potential to provide increased accuracy and reliability across a wide range of applications. The results indicate that the proposed models outperform existing models in terms of accuracy and robustness for all three datasets. The source code can be found at https://huggingface.co/DebajyotyBanik/Ensemble-Pipelined-Regression/tree/main

The astrophysical parameters of chemically peculiar stars from automatic methods

Context. The chemically peculiar (CP) stars of the upper main sequence are excellent astrophysical laboratories for investigating the diffusion, mass loss, rotational mixing, and pulsation in the presence and absence of a stable local magnetic field. For this, we need a homogeneous set of parameters, such as effective temperature (Teff) and surface gravity (log g), to locate the stars in the Hertzsprung-Russell diagram so that we can then estimate the mass, radius, and age. Aims. In recent years, the results of several automatic pipelines have been published; these use various techniques and data sets, including Teff and log g values for millions of stars. Because CP stars are known to have flux anomalies, these astrophysical parameters must be tested for their reliability and usefulness. If the outcome is positive, these can be used to analyse the new and faint CP stars published recently. Methods. I compared published Teff and log g values of a set of CP stars, which are mostly based on high-resolution spectroscopy, with values from four automatic pipeline approaches. In doing so, I searched for possible correlations and offsets. Results. I present a detailed statistical analysis of a comparison between the ‘standard’ and published Teff and log g values. The accuracy depends on the presence of a magnetic field and the spectral type of the CP subgroups. However, I obtain standard deviations of between 2% and 20%. Conclusions. Considering the statistical errors, the astrophysical parameters from the literature can be used for CP stars, although caution is advised for magnetic CP stars.

Bioinformatics and Systems Biology Approaches to Identify the Synergistic Effects of Alcohol Use Disorder on the Progression of Neurological Diseases

Clinical investigations showed that individuals with Alcohol Use Disorder (AUD) have worse Neurological Disease (ND) development, pointing to possible pathogenic relationships between AUD and NDs. It remains difficult to identify risk factors that are predisposing between AUD and NDs. In order to fix these issues, we created the bioinformatics pipeline and network-based approaches for employing unbiased methods to discover genes abnormally stated in both AUD and NDs and to pinpoint some of the common molecular pathways that might underlie AUD and ND interaction. We found 100 differentially expressed genes (DEGs) in both the AUD and ND patient’s tissue samples. The most important Gene Ontology (GO) terms and metabolic pathways, including positive control of cytotoxicity caused by T cells, proinflammatory responses, antigen processing and presentation, and platelet-triggered interactions with vascular and circulating cell pathways were then extracted using the overlapped DEGs. Protein-protein interaction analysis was used to identify hub proteins, including CCL2, IL1B, TH, MYCN, HLA-DRB1, SLC17A7, and HNF4A, in the pathways that have been reported as playing a function in these disorders. We determined several TFs (HNF4A, C4A, HLA-B, SNCA, HLA-DMB, SLC17A7, HLA-DRB1, HLA-C, HLA-A, and HLA-DPB1) and potential miRNAs (hsa-mir-34a-5p, hsa-mir-34c-5p, hsa-mir-449a, hsa-mir-155-5p, and hsa-mir-1-3p) were crucial for regulating the expression of AUD and ND which could serve as prospective targets for treatment. Our methodologies discovered unique putative biomarkers that point to the interaction between AUD and various neurological disorders, as well as pathways that could one day be the focus of therapeutic intervention.

Identification of Potentially Repurposable Drugs for Lewy Body Dementia Using a Network-Based Approach.

The conventional method of one drug being used for one target has not yielded therapeutic solutions for Lewy body dementia (LBD), which is a leading progressive neurological disorder characterized by significant loss of neurons. The age-related disease is marked by memory loss, hallucinations, sleep disorder, mental health deterioration, palsy, and cognitive impairment, all of which have no known effective cure. The present study deploys a network medicine pipeline to repurpose drugs having considerable effect on the genes and proteins related to the diseases of interest. We utilized the novel SAveRUNNER algorithm to quantify the proximity of all drugs obtained from DrugBank with the disease associated gene dataset obtained from Phenopedia and targets in the human interactome. We found that most of the 154 FDA-approved drugs predicted by SAveRUNNER were used to treat nervous system disorders, but some off-label drugs like quinapril and selegiline were interestingly used to treat hypertension and Parkinson's disease (PD), respectively. Additionally, we performed gene set enrichment analysis using Connectivity Map (CMap) and pathway enrichment analysis using EnrichR to validate the efficacy of the drug candidates obtained from the pipeline approach. The investigation enabled us to identify the significant role of the synaptic vesicle pathway in our disease and accordingly finalize 8 suitable antidepressant drugs from the 154 drugs initially predicted by SAveRUNNER. These potential anti-LBD drugs are either selective or non-selective inhibitors of serotonin, dopamine, and norepinephrine transporters. The validated selective serotonin and norepinephrine inhibitors like milnacipran, protriptyline, and venlafaxine are predicted to manage LBD along with the affecting symptomatic issues.

Biomedical event extraction using pre-trained SciBERT

Abstract Biomedical event extraction is applied to biomedical texts to obtain a list of events within the biomedical domain. The best GENIA biomedical event extraction research uses sequence labeling techniques with a joint approach, softmax decoder for event trigger identification, and the BioBERT v1.1 encoder. However, this event extraction model has three drawbacks: tasks are carried out independently, it does not provide special handling of multi-label event trigger labels, and it uses an encoder with vocabulary from non-biomedical domains. We propose to use the pipeline approach to provide forward information sigmoid to address multi-label event trigger labels and alternative BERT encoders with vocabulary from the biomedical domain. The experiment showed that the performance of the biomedical event extraction model increased after changing the encoder, which had been built using a biomedical-specific domain vocabulary. Changing the encoder to SciBERT while still using the joint approach and softmax decoder increased the precision by 4.22 points (reaching 69.88) and resulted in an F1-score of 58.48.

A pipeline approach for privacy preservation against poisoning attacks in a Mobile Edge Computing environment

Mobile Edge Computing (MEC) has become increasingly popular due to its ability to provide real-time data processing and improved user experience. However, it also introduces potential security vulnerabilities, as malicious actors could exploit users’ position data stored in the MEC server. To address this challenge, it is essential to develop privacy-preserving schemes that are resistant to malicious activities, such as poisoning attacks. Previous works have not considered the evaluation of such schemes against certain types of poisoning attacks while ensuring sufficient utility, privacy, and scalability during the generation process of poisoning locations. Indeed, meeting the needs of the user or customers, keeping data secure and confidential, and having capacity to accommodate changes in demand are essential for any application or system to operate efficiently and securely. To address this limitation, we present a new pipeline approach for a privacy-preserving scheme against poisoning attacks in the MEC environment. We evaluate this method with a type of poisoning attack that ensures sufficient utility, privacy, and scalability during the generation process and assess its effectiveness in providing security and privacy. Our approach is evaluated on two data sets and the results are compared with three baseline approaches. The results show that our approach is more effective in detecting poisoning locations than the existing works without affecting the QoS metrics (bandwidth, resources consumption, latency and the storage) with should be consider when evaluating MEC method.

LOW-COST CLOUD-BASED HD-MAP UPDATES FOR INFRASTRUCTURE MANAGEMENT AND MAINTENANCE

Abstract. Recently, HD maps have various merits for achieving the highest level of self-localization accuracy, keeping track of the state of the road infrastructure and maintenance, and providing an indication if any repairs are required. Therefore, it is essential to keep the HD map up to date. However, the process of updating the HD map is exorbitant because the HD map is created using expensive sensor setups, and updating the map frequently via these setups will be costly. In this paper, a full pipeline is proposed for updating the HD map via a crowdsourced dataset that is collected with low-cost smartphone sensors. Furthermore, an Android application is developed and installed on a smartphone to collect the raw data. Once the dataset is collected from the area of interest, it will be uploaded automatically to the cloud server that is connected to the HD map database. Then, object detection, depth estimation, and matching algorithms are triggered on the cloud server to keep updating the HD map database. The positions of the detected objects from the crowdsourced dataset are estimated by using fused outputs of deep learning models and the Global navigation satellite system (GNSS) of a smartphone and then compared with the objects in the HD map through matching algorithms. The proposed model is considered the first comprehensive pipeline approach for updating HD maps with high a cost-effective and efficient solution.

Pipelined correlated minimum weight perfect matching of the surface code

We describe a pipeline approach to decoding the surface code using minimum weight perfect matching, including taking into account correlations between detection events. An independent no-communication parallelizable processing stage reweights the graph according to likely correlations, followed by another no-communication parallelizable stage for high confidence matching. A later general stage finishes the matching. This is a simplification of previous correlated matching techniques which required a complex interaction between general matching and re-weighting the graph. Despite this simplification, which gives correlated matching a better chance of achieving real-time processing, we find the logical error rate practically unchanged. We validate the new algorithm on the fully fault-tolerant toric, unrotated, and rotated surface codes, all with standard depolarizing noise. We expect these techniques to be applicable to a wide range of other decoders.

Decision rules for personalized statin treatment prescriptions over multi-objectives.

In our previous study, we demonstrated the feasibility of producing a proactive statin prescription strategy - a personalized statin treatment plan (PSTP) - using neural networks with big data. However, its non-transparency limited result interpretations and clinical usability. To improve the transparency of our previous approach with minimal compromise to the maximal statin treatment benefit-to-risk ratio, this study proposed a five-step pipeline approach called the decision rules for statin treatment (DRST). Steps 1-3 of our proposed pipeline improved our previous PSTP model in optimizing individual benefit-to-risk ratio; Step 4 used a decision tree model (DRST) to provide straightforward rules in the initial statin treatment plan; Step 5 aimed to evaluate the efficacy of these decision rules by conducting a clinical trial simulation. We included 107,739 de-identified patient data from Optum Labs Database Warehouse in this study. The final decision rules were compact and efficient, resulting from a decision tree with only a maximum depth of 3 and 11 nodes. The DRST identified three factors that are easily obtainable at the point of care: age, low-density lipoprotein cholesterol (LDL-C) level, and age-adjusted Charlson score. Moreover, it also identified six subpopulations that can benefit most from these decision rules. In our clinical trial simulations, DRST was found to improve statin benefit in LDL-C reduction by 4.15 percentage points (pp) and reduce risks of statin-associated symptoms (SAS) and statin discontinuation by 11.71 and 3.96 pp, respectively, when compared to the standard of care. Moreover, these DRST results were only less than 0.6 pp suboptimal to PSTP, demonstrating that building DRST that provide transparency with minimal compromise to the maximal benefit-to-risk ratio of statin treatments is feasible.

Computational Pipeline for Analysis of Biomedical Networks with BioNAR.

In a living cell, proteins interact to assemble both transient and constant molecular complexes, which transfer signals/information around internal pathways. Modern proteomic techniques can identify the constituent components of these complexes, but more detailed analysis demands a network approach linking the molecules together and analyzing the emergent architectural properties. The Bioconductor package BioNAR combines a selection of existing R protocols for network analysis with newly designed original methodological features to support step-by-step analysis of biological/biomedical . Critically, BioNAR supports a pipeline approach whereby many networks and iterative analyses can be performed. Here we present a network analysis pipeline that starts from initiating a network model from a list of components/proteins and their interactions through to identifying its functional components based solely on network topology. We demonstrate that BioNAR can help users achieve a number of network analysis goals that are difficult to achieve anywhere else. This includes how users can choose the optimal clustering algorithm from a range of options based on independent annotation enrichment, and predict a protein's influence within and across multiple subcomplexes in the network and estimate the co-occurrence or linkage between metadata at the network level (e.g., diseases and functions across the network, identifying the clusters whose components are likely to share common function and mechanisms). The package is freely available in Bioconductor release 3.17: https://bioconductor.org/packages/3.17/bioc/html/BioNAR.html. © 2023 The Authors. Current Protocols published by Wiley Periodicals LLC. Basic Protocol 1: Creating and annotating the network Support Protocol 1: Installing BioNAR from RStudio Support Protocol 2: Building the sample network from synaptome.db Basic Protocol 2: Network properties and centrality Basic Protocol 3: Network communities Basic protocol 4: Choosing the optimal clustering algorithm based on the enrichment with annotation terms Basic Protocol 5: Influencing network components and bridgeness Basic Protocol 6: Co-occurrence of the annotations.

Antioxidants green tea extract and nordihydroguaiaretic acid confer species and strain-specific lifespan and health effects in Caenorhabditis nematodes

The Caenorhabditis Intervention Testing Program (CITP) is an NIH-funded research consortium of investigators who conduct analyses at three independent sites to identify chemical interventions that reproducibly promote health and lifespan in a robust manner. The founding principle of the CITP is that compounds with positive effects across a genetically diverse panel of Caenorhabditis species and strains are likely engaging conserved biochemical pathways to exert their effects. As such, interventions that are broadly efficacious might be considered prominent compounds for translation for pre-clinical research and human clinical applications. Here, we report results generated using a recently streamlined pipeline approach for the evaluation of the effects of chemical compounds on lifespan and health. We studied five compounds previously shown to extend C. elegans lifespan or thought to promote mammalian health: 17α-estradiol, acarbose, green tea extract, nordihydroguaiaretic acid, and rapamycin. We found that green tea extract and nordihydroguaiaretic acid extend Caenorhabditis lifespan in a species-specific manner. Additionally, these two antioxidants conferred assay-specific effects in some studies—for example, decreasing survival for certain genetic backgrounds in manual survival assays in contrast with extended lifespan as assayed using automated C. elegans Lifespan Machines. We also observed that GTE and NDGA impact on older adult mobility capacity is dependent on genetic background, and that GTE reduces oxidative stress resistance in some Caenorhabditis strains. Overall, our analysis of the five compounds supports the general idea that genetic background and assay type can influence lifespan and health effects of compounds, and underscores that lifespan and health can be uncoupled by chemical interventions.

Piezoelectric nanogenerators for self-powered wearable and implantable bioelectronic devices.

One of the recent innovations in the field of personalized healthcare is the piezoelectric nanogenerators (PENGs) for various clinical applications, including self-powered sensors, drug delivery, tissue regeneration etc. Such innovations are perceived to potentially address some of the unmet clinical needs, e.g., limited life-span of implantable biomedical devices (e.g., pacemaker) and replacement related complications. To this end, the generation of green energy from biomechanical sources for wearable and implantable bioelectronic devices gained considerable attention in the scientific community. In this perspective, this article provides a comprehensive state-of-the-art review on the recent developments in the processing, applications and associated concerns of piezoelectric materials (synthetic/biological) for personalized healthcare applications. In particular, this review briefly discusses the concepts of piezoelectric energy harvesting, piezoelectric materials (ceramics, polymers, nature-inspired), and the various applications of piezoelectric nanogenerators, such as, self-powered sensors, self-powered pacemakers, deep brain stimulators etc. Important distinction has been made in terms of the potential clinical applications of PENGs, either as wearable or implantable bioelectronic devices. While discussing the potential applications as implantable devices, the biocompatibility of the several hybrid devices using large animal models is summarized. This review closes with the futuristic vision of integrating data science approaches in developmental pipeline of PENGs as well as clinical translation of the next generation PENGs. STATEMENT OF SIGNIFICANCE: Piezoelectric nanogenerators (PENGs) hold great promise for transforming personalized healthcare through self-powered sensors, drug delivery systems, and tissue regeneration. The limited battery life of implantable devices like pacemakers presents a significant challenge, leading to complications from repititive surgeries. To address such a critical issue, researchers are focusing on generating green energy from biomechanical sources to power wearable and implantable bioelectronic devices. This comprehensive review critically examines the latest advancements in synthetic and nature-inspired piezoelectric materials for PENGs in personalized healthcare. Moreover, it discusses the potential of piezoelectric materials and data science approaches to enhance the efficiency and reliability of personalized healthcare devices for clinical applications.

Accelerating ray tracing engine of BLENDER on the new Sunway architecture

AbstractWith the increasing popularity of high‐resolution displays, there is a growing demand for more realistic rendered images. Ray tracing has become the most effective algorithm for image rendering, but its complexity and large amount of computing data require sophisticated HPC solutions. In this article, we present our efforts to port the ray tracing engine CYCLES of Blender to the new generation of Sunway supercomputers. We propose optimizations that are tailored to the new hardware architecture, including a multi‐level parallel scheme that efficiently maps and scales Blender onto the novel Sunway architecture, strategies to address memory bottlenecks, a revised task dispatching method that achieves excellent load balancing, and a pipeline approach that maximizes computation and communication overlap. By combining all these optimizations, we achieve a significant reduction in rendering time for a single‐frame image, from 2260 s using the single‐core serial version to 71 s using 48 processes, which is a speedup of about 128×. Accelerating the ray tracing engine CYCLES of Blender in the new generation of Sunway supercomputers.

EI-RNN-based text generation for the static and dynamic isolated sign language videos

SLR, which assists hearing-impaired people to communicate with other persons by sign language, is considered as a promising method. However, as the features of some of the static SL could be the same as the feature in a single frame of dynamic Isolated Sign Language (ISL), the generation of accurate text corresponding to the SL is necessary during the SLR. Therefore, Edge-directed Interpolation-based Recurrent Neural Network (EI-RNN)-centered text generation with varied features of the static and dynamic Isolated SL is proposed in this article. Primarily, ISL videos are converted to frames and pre-processed with key frame extraction and illumination control. After that, the foreground is separated with the Symmetric Normalised Laplacian-centered Otsu Thresholding (SLOT) technique for finding accurate key points in the human pose. The human pose’s key points are extracted with the Media Pipeline Holistic (MPH) pipeline approach and to improve the features of the face and hand sign, the resultant frame is fused with the depth image. After that, to differentiate the static and dynamic actions, the action change in the fused frames is determined with a correlation matrix. After that, to engender the output text for the respective SL, features are extracted individually as of the static and dynamic frames. It is obtained from the analysis that when analogized to the prevailing models, the proposed EI-RNN’s translation accuracy is elevated by 2.05% in INCLUDE 50 Indian SL based Dataset and Top 1 Accuracy 2.44% and Top 10 accuracy, 1.71% improved in WLASL 100 American SL.

Automatic peak annotation and area estimation of glycan map peaks directly from chromatograms

AbstractThe present bottleneck in biosimilar bioprocess development has become evaluation of analytical results, due to recent advances in analytics, such as automated sample preparation and development of high‐throughput methods. Currently automated chromatogram integration and annotation is only efficient for simple chromatograms. In an ever more competitive field of biosimilars, this represents a serious drawback because chromatographic analytical methods that provide some of the most valuable physicochemical quality attributes of the product also require careful chromatogram integration and annotation. This work focuses on the glycan mapping analytical method as utilized in development of monoclonal antibody biosimilars, evaluating more than 2000 chromatograms spanning the life cycle of multiple biosimilar development projects. It proposes a modified workflow by implementing automatic machine learning algorithms to determine the proportion of specific relevant glycan species in a sample directly from the chromatogram. Data preparation and analysis is performed using a pipeline approach. Pipeline is a modular design of data processing where signal “travels” through various active modules in a series. Each module performs a specific function or transformation on the signal and propagates the transformed signal to the next module. The pipeline is designed in a way that modules can be independently improved and exchanged. Module functions currently implemented are chromatogram resampling by spline interpolation, baseline removal by asymmetric least squares, peak alignment using parametric time warping, and quantification of the relative proportion of a glycan species using partial least squares regression. Hyper‐parameters of the pipeline are then optimized using the Nelder–Mead method. The approach stands out for its ability to accommodate a broad landscape of samples, covering multiple different proteins in different stages of biosimilar development, analyzed using different adaptations of the glycan map analytical method. The pipeline presents an intuitive, flexible, and creatively simple method design capable of providing reliable results for a wide range of glycan species essential for biosimilar development. It enables transparent, faster, and less subjective evaluation of analytic raw data (from sample to result). Furthermore, our automated approach maintained an accuracy comparable with manual integration thus demonstrating its readiness for implementation in the conservative and highly regulated environment. The presented methodology reduces the cost and time of biosimilar development and should be applicable for any chromatogram‐based analytical method.

A Workflow Combining Machine Learning with Molecular Simulations Uncovers Potential Dual-Target Inhibitors against BTK and JAK3.

The drug development process suffers from low success rates and requires expensive and time-consuming procedures. The traditional one drug-one target paradigm is often inadequate to treat multifactorial diseases. Multitarget drugs may potentially address problems such as adverse reactions to drugs. With the aim to discover a multitarget potential inhibitor for B-cell lymphoma treatment, herein, we developed a general pipeline combining machine learning, the interpretable model SHapley Additive exPlanation (SHAP), and molecular dynamics simulations to predict active compounds and fragments. Bruton's tyrosine kinase (BTK) and Janus kinase 3 (JAK3) are popular synergistic targets for B-cell lymphoma. We used this pipeline approach to identify prospective potential dual inhibitors from a natural product database and screened three candidate inhibitors with acceptable drug absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties. Ultimately, the compound CNP0266747 with specialized binding conformations that exhibited potential binding free energy against BTK and JAK3 was selected as the optimum choice. Furthermore, we also identified key residues and fingerprint features of this dual-target inhibitor of BTK and JAK3.