Open Source Web Research Articles

New algorithms demonstrate untargeted detection of chemically meaningful changing units and formula assignment for HRMS data of polymeric mixtures in the open-source constellation web application

The field of high-resolution mass spectrometry (HRMS) and ancillary hyphenated techniques comprise a rapidly expanding and evolving area. As popularity of HRMS instruments grows, there is a concurrent need for tools and solutions to simplify and automate the processing of the large and complex datasets that result from these analyses. Constellation is one such of these tools, developed by our group over the last two years to perform unsupervised trend detection for repeating, polymeric units in HRMS data of complex mixtures such as natural organic matter, oil, or lignin. In this work, we develop two new unsupervised algorithms for finding chemically-meaningful changing units in HRMS data, and incorporate a molecular-formula-finding algorithm from the open-source CoreMS software package, both demonstrated here in the Constellation software environment. These algorithms are evaluated on a collection of open-source HRMS datasets containing polymeric analytes (PEG 400 and NIST standard reference material 1950, both metabolites in human plasma, as well as a swab extract containing polymers), and are able to successfully identify all known changing units in the data, including assigning the correct formulas. Through these new developments, we are excited to add to a growing body of open-source software specialized in extracting useful information from complex datasets without the high costs, technical knowledge, and processor-demand typically associated with such tools.

TraceTrack, an open-source software for batch processing, alignment and visualization of sanger sequencing chromatograms.

Despite the advent of next-generation sequencing technology and its widespread applications, Sanger sequencing remains instrumental for molecular biology subcloning work in biological and medical research and indispensable for drug discovery campaigns. Although Sanger sequencing technology has been long established, existing software for processing and visualization of trace file chromatograms is limited in terms of functionality, scalability and availability for commercial use. To fill this gap, we developed TraceTrack, an open-source web application tool for batch alignment, analysis and visualization of Sanger trace files. TraceTrack offers high-throughput matching of trace files to reference sequences, rapid identification of mutations and an intuitive chromatogram analysis. Comparative analysis between TraceTrack and existing software tools highlights the advantages of TraceTrack with regards to batch processing, visualization and export functionalities. TraceTrack is available at https://github.com/MSDLLCpapers/TraceTrack and as a web application at https://tracetrack.dichlab.org. TraceTrack is a web application for batch processing and visualization of Sanger trace file chromatograms that meets the increasing demand of industrial sequence validation workflows in pharmaceutical settings. Supplementary data are available at Bioinformatics Advances online.

Investigating unique genes of five molecular subtypes of breast cancer using penalized logistic regression.

Breast cancer (BC) is the most common cancer and the fifth cause of death in women worldwide. Exploring unique genes for cancers has been interesting. This study aimed to explore unique genes of five molecular subtypes of BC in women using penalized logistic regression models. For this purpose, microarray data of five independent GEO data sets were combined. This combination includes genetic information of 324 women with BC and 12 healthy women. Least absolute shrinkage and selection operator (LASSO) logistic regression and adaptive LASSO logistic regression were used to extract unique genes. The biological process of extracted genes was evaluated in an open-source GOnet web application. R software version 3.6.0 with the glmnet package was used for fitting the models. Totally, 119 genes were extracted among 15 pairwise comparisons. Seventeen genes (14%) showed overlap between comparative groups. According to GO enrichment analysis, the biological process of extracted genes was enriched in negative and positive regulation biological processes, and molecular function tracking revealed that most genes are involved in kinase and transferring activities. On the other hand, we identified unique genes for each comparative group and the subsequent pathways for them. However, a significant pathway was not identified for genes in normal-like versus ERBB2 and luminal A, basal versus control, and lumina B versus luminal A groups. Most genes selected by LASSO logistic regression and adaptive LASSO logistic regression identified unique genes and related pathways for comparative subgroups of BC, which would be useful to comprehend the molecular differences between subgroups that would be considered for further research and therapeutic approaches in the future.

Prediction of Continuous Rainfall-Runoff Scenarios for Snow-Fed Watershed of Himachal Pradesh, India using Geographical Information System GIS based on Hydrological Modelling

Rainfall runoff modeling is one of the most complex hydrological modeling due to the involvement of different watershed physical parameters. It is essential for the analysis of watershed hydrological response towards the received precipitation under the influence of watershed variables. As it is a replica of watershed hydrological response, Rainfall-Runoff modeling is essential to evaluate the general characteristics of total surface runoff at catchment’s outlet. The main objective of this study was prediction of rainfall runoff using Hydrologic Engineering Center-Hydrologic Modeling System [HEC-HMS] model for Rampur watershed which was covered in Kullu and Shimla districts situated at Upper Sutlej sub-basin of Himachal Pradesh, India. Geo-spatial data of Digital Elevation Model,LandUse-LandCover (LU-LC) map, Soil map and Hydro-meteorological data of rainfall were collected from different open source web portals to utilise as input parameters for modeling. Soil Conservation Service-Curve Number (SCS-CN) for precipitation loss modeling, Soil Conservation Service–Dimensionless Unit Hydrograph [SCS-UH] for transform modeling, Muskingum method for Runoff routing were used as input parameters of respective components of HEC-HMS model. This article proposes a methodology for generating hourly rainfall data from daily rainfall data using Intensity-Duration-Frequencycurves in Meteorological model of HMS. The simulation results of runoff with 5 different return period scenarios are follows the return period frequency tendency, such as 1000 year return period runoff peak delivers 4729.5 m3/s, 500 year return period delivers a peak runoff of 4263.9 m3/s, 100 year return period gives a peak runoff of 3196.5 m3/s, 50 year return period runoff deliver 2745.3 m3/s and 10 year return period deliver 953.7 m3/s runoff at the outlet of Rampur watershed.

Sanare: Pluggable Intrusion Recovery for Web Applications

Web applications are exposed to many threats and, despite the best defensive efforts, are often successfully attacked. Reverting the effects of an attack on the state of such an application requires a profound knowledge about the application, to understand what data did the attack corrupt. Furthermore, it requires knowing what steps are needed to revert the effects without modifying legitimate data created by legitimate users. Existing intrusion recovery systems are capable of reverting the effects of the attack but they require modifications to the source code of the application, which may be unpractical. We present <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Sanare</i> , a pluggable intrusion recovery system designed for web applications that use different data storage systems to keep their state. Sanare does not require any modification to the source code of the application or the web server. Instead, it uses a new deep learning scheme that we also introduce in the article, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Matchare</i> , that learns the matches between the HTTP requests and the database statements, file system operations, and web service requests that the HTTP requests caused. We evaluated Sanare with three open source web applications: WordPress, GitLab and ownCloud. In our experiments, Matchare achieved precision and recall higher than 97.5% with a performance overhead of less than 18% to the application.

Rianú: Multi-tissue tracking software for increased throughput of engineered cardiac tissue screening

Background:The field of tissue engineering has provided valuable three-dimensional species-specific models of the human myocardium in the form of human Engineered Cardiac Tissues (hECTs) and similar constructs. However, hECT systems are often bottlenecked by a lack of openly available software that can collect data from multiple tissues at a time, even in multi-tissue bioreactors, which limits throughput in phenotypic and therapeutic screening applications. Methods:We developed Rianú, an open-source web application capable of simultaneously tracking multiple hECTs on flexible end-posts. This software is operating system agnostic and deployable on a remote server, accessible via a web browser with no local hardware or software requirements. The software incorporates object-tracking capabilities for multiple objects simultaneously, an algorithm for twitch tracing analysis and contractile force calculation, and a data compilation system for comparative analysis within and amongst groups. Validation tests were performed using in-silico and in-vitro experiments for comparison with established methods and interventions. Results:Rianú was able to detect the displacement of the flexible end-posts with a sub-pixel sensitivity of 0.555 px/post (minimum increment in post displacement) and a lower limit of 1.665 px/post (minimum post displacement). Compared to our established reference for contractility assessment, Rianú had a high correlation for all parameters analyzed (ranging from R2=0.7514 to R2=0.9695), demonstrating its high accuracy and reliability. Conclusions:Rianú provides simultaneous tracking of multiple hECTs, expediting the recording and analysis processes, and simplifies time-based intervention studies. It also allows data collection from different formats and has scale-up capabilities proportional to the number of tissues per field of view. These capabilities will enhance throughput of hECTs and similar assays for in-vitro analysis in disease modeling and drug screening applications.

A web scraping-based approach for fundamental analysis platform in financial assets

There are two main benefits of using fundamental analysis for investors and portfolio managers. First, investing in a company with good ratios has lower risks. The second reason is that it is possible to evaluate share prices with internal valuation methods based on ratios. These price valuations can be more meaningful when combined with technical analysis data. Many data terminals provide processes such as fundamental analysis data and price valuation on a paid and licensed basis. However, the balance sheet data of publicly traded markets are publicly available and can be obtained and interpreted by web scraping methods. This study presents an approach in which basic analysis and price evaluation are made with balance sheets and ratios using open-source tools and web scraping.

COVID19PREDICTOR: KLİNİK VERİLERE VE RUTİN TESTLERE DAYALI OLARAK COVID-19 TEŞHİSİ İÇİN MAKİNE ÖĞRENİMİ MODELLERİ GELİŞTİRMEYE YARAYAN WEB TABANLI ARAYÜZ

Objective: The Covid-19 outbreak has become the primary health problem of many countries due to health related, social, economic and individual effects. In addition to the development of outbreak prediction models, the examination of risk factors of the disease and the development of models for diagnosis are of high importance. This study introduces the Covid19PredictoR interface, a workflow where machine learning approaches are used for diagnosing Covid-19 based on clinical data such as routine laboratory test results, risk factors, information on co-existing health conditions.&#x0D; Method: Covid19PredictoR interface is an open source web based interface on R/Shiny (https://biodatalab.shinyapps.io/Covid19PredictoR/). Logistic regression, C5.0, decision tree, random forest and XGBoost models can be developed within the framework. These models can also be used for predictive purposes. Descriptive statistics, data pre-processing and model tuning steps are additionally provided during model development.&#x0D; Results: Einsteindata4u dataset was analyzed with the Covid19PredictoR interface. With this example, the complete operation of the interface and the demonstration of all steps of the workflow have been shown. High performance machine learning models were developed for the dataset and the best models were used for prediction. Analysis and visualization of features (age, admission data and laboratory tests) were carried out for the case per model.&#x0D; Conclusion: The use of machine learning algorithms to evaluate Covid-19 disease in terms of related risk factors is rapidly increasing. The application of these algorithms on various platforms creates application difficulties, repeatability and reproducibility problems. The proposed pipeline, which has been transformed into a standard workflow with the interface, offers a user-friendly structure that healthcare professionals with various background can easily use and report.

Linked Open Images: Visual similarity for the Semantic Web

This paper presents ArtVision, a Semantic Web application that integrates computer vision APIs with the ResearchSpace platform, allowing for the matching of similar artworks and photographs across cultural heritage image collections. The field of Digital Art History stands to benefit a great deal from computer vision, as numerous projects have already made good progress in tackling issues of visual similarity, artwork classification, style detection, gesture analysis, among others. Pharos, the International Consortium of Photo Archives, is building its platform using the ResearchSpace knowledge system, an open-source semantic web platform that allows heritage institutions to publish and enrich collections as Linked Open Data through the CIDOC-CRM, and other ontologies. Using the images and artwork data of Pharos collections, this paper outlines the methodologies used to integrate visual similarity data from a number of computer vision APIs, allowing users to discover similar artworks and generate canonical URIs for each artwork.

Jot: guiding journal selection with suitability metrics.

Researchers grapple with a challenging and consequential decision each time they choose a journal for manuscript submission. There are several online tools that attempt to identify appropriate journals for a manuscript, but each of these tools has shortcomings in terms of the journal data they provide and the exploration functionality they offer-and not one of these tools is open source. Jot is a free and open-source web application that matches manuscripts in the fields of biomedicine and life sciences with suitable journals, based on a manuscript's title, abstract, and (optionally) citations. Jot gathers a wealth of data on journal quality, impact, fit, and open access options that can be explored through a dashboard of linked, interactive visualizations.

OSADHI – An online structural and analytics based database for herbs of India

The current study aims to develop a PAN India database of medicinal plants along with their phytochemicals and geographical availability. The database consists of 6959 unique medicinal plants belonging to 348 families which are available across 28 states and 8 union territories of India. The database sources the information on four different sections – traditional knowledge, geographical indications, phytochemicals, and chemoinformatics. The traditional knowledge reports the plant taxonomy with their vernacular names. A total of 27,440 unique phytochemicals associated with these plants were curated from various sources in this study. However, due to the non-availability of general information like IUPAC names, InChI key, etc. from reliable sources, only 22,314 phytochemicals have been currently reported in the database. Various analyses have been performed for the phytochemicals which include analysis of physicochemical and ADMET properties calculated from open-source web servers using in-house python scripts. The phytochemical data set has also been classified based on the class, superclass, and pathways respectively using NPClassifier, a deep learning framework. Additionally, the antiviral potency of the phytochemicals was also predicted using two machine learning models – Random Forest and XGBoost. The database aims to provide accurate and exhaustive data of the traditional practice of medicinal plants in India in a single platform integrating and analyzing the rich customary practices and facilitating the development and identification of plant-based therapeutics for a variety of diseases. The database can be accessed at https://neist.res.in/osadhi/.

From Unit to Dose: A Machine Learning Approach for Precise Prediction of Hemoglobin and Iron Content in Individual Packed Red Blood Cell Units.

Transfusion of packed red blood cells (pRBCs) saves lives, but iron overload limits survival of chronically transfused patients. Quality control methods, which involve entering pRBC units and removing them from the blood supply, reveal that hemoglobin (38.5-79.9 g) and heme iron (133.42-276.89 mg) vary substantially between pRBCs. Yet, neither hemoglobin nor iron content can be quantified for individual clinically used pRBCs leading to rules of thumb for pRBC transfusions. Keeping their integrity, the authors seek to predict hemoglobin/iron content of any given pRBC unit applying eight machine learning models on 6,058 pRBCs. Based on thirteen features routinely collected during blood donation, production and quality control testing, the model with best trade-off between performance and complexity in hemoglobin/iron content prediction is identified. Validation of this model in an independent cohort of 2637 pRBCs confirms an adjusted R2 > 0.9 corresponding to a mean absolute prediction error of ≤1.43 g hemoglobin/4.96 mg iron (associated standard deviation: ≤1.13 g hemoglobin/3.92 mg iron). Such unprecedented precise prediction enables reliable pRBC dosing per pharmaceutically active agent, and monitoring iron uptake in patients and individual iron loss in donors. The model is implemented in a free open source web application to facilitate clinical application.

54. CancerTrialMatch: A computational resource for the management of clinical trials at a precision oncology center

The routine use of genetic and genomic sequencing in the treatment of cancer has been steadily rising over the past few years, with a corresponding increase in the availability of biomarker-based clinical trials. At the Avera Cancer Institute, biomarker-based clinical trials are often presented as treatment options to oncologists at the molecular tumor board. This necessitated a way to capture structured trial data, and match them to patients based on their disease and sequencing profile in a systematic manner. We developed an open-source web application, CancerTrialMatch, that has the ability to (i) add trials through a semi-automated curation interface, (ii) browse and search trials, (iii) and match patients to biomarker-based trials. This application uses R Shiny to create simple interfaces, a mongo database to store trial data, various R libraries to query data and perform computations, and Docker to manage software installation and application instantiation. The curation interface is semi-automated because querying the clinicaltrials.gov API will return discrete data for many fields, but not for biomarkers or for disease subtypes. The user has to manually input disease type based on the OncoTree classification, as well as biomarker information for mutations, copy numbers, fusions, TMB, MSI/PD-L1 status, RNA expression, and disease-specific markers such as ER/PR/HER2 status. We believe that this application will reduce the person-hours required for trial management for a patient, aid in increasing clinical trial enrollment by systematically providing treatment options, and is thus an important tool in the clinical application of precision oncology.

THREE-DIMENSIONAL ENABLEMENT OF PLACE-BASED, PANDEMIC BEHAVIORS

Abstract. Harvesting usable and meaningful disaster-related, spatio-temporal data at a highly granular level poses major challenges in its cleaning and aggregation. This paper presents a strategy related to those challenges with respect to individual behavior near COVID-19 laden healthcare facilities. This is done to enable the visualizing of egress behavior data as interactive, three-dimensional (3D) scenes to investigate human behavior patterns regarding touch-based, disease transmission. Therefore, the aim is to demonstrate how this concept of 3D epidemiology may provide new mechanisms to understand the relative risk and exposure prevalence for data analysis. This paper demonstrates 3D enablement of disaster-related field data through use of first-hand observations of 1,936 individuals egressing New York City healthcare facilities during the onset of COVID-19 in the Spring of 2020. The observations capture egress behavior in terms of where people go (e.g. coffee shop, Subway) and how they physically interact with the surroundings (i.e. what they touch and how long they remain). This paper introduces a mechanism for automated extraction and 3D visualization of such data in Potree, an open-source Web Graphics Library (WebGL) point cloud viewer. Distinctive vertex shaders are used to distinguish specific destination selection and behavioral patterns (e.g. personal protective equipment usage). Two-dimensional heatmaps are paired with 3D scenes to demonstrate the potential of using 3D visualization of spatio-temporal patterns for visualizing disease transmission potential.

A machine learning analysis of patient concerns regarding mastopexy

Social media plays an important role in connecting patients and plastic surgeons. We utilized patient inquiries regarding mastopexy from an online social media site to determine the most prevalent patient concerns, while employing a machine-learning algorithm to generate the questions representative of the dataset. This data allow plastic surgeons to better tailor their preoperative consultations to address common concerns, set realistic expectations, and improve overall satisfaction. A total of 2,011 inquiries from the mastopexy section of Realself.com were obtained using an open-source web crawler. Each inquiry was manually categorized as preoperative or postoperative and classified into subcategories based upon the free text entry. Lastly, questions were analyzed using machine learning to determine ten questions most representative of the inquiry pool. Of the 2,011 inquiries analyzed, 52.91% were preoperative and 47.09% were postoperative. Most preoperative questions asked about procedure eligibility (309, 29.04%), surgical techniques and logistics (260, 24.44%), and the best type of breast lift for the user (259, 24.34%). Among postoperative questions, questions regarding appearance were the most common (491, 51.85%), followed by symptoms after surgery (197, 19.75%) and behavior allowed/disallowed (145, 15.31%). Appearance was further subcategorized with the most common categories being appearance of the nipple (98, 19.86%), skin discoloration (88, 17.92%), and scarring (74, 15.07%). By utilizing the data that social media websites, like Realself.com, provide, plastic surgeons can better understand common patient concerns. This data aid in optimizing the preoperative consultation process to address the common concerns, recalibrate unrealistic expectations, and improve overall satisfaction.

Rom.js/cfd.xyz: An open-source framework for generating and visualizing parametric CFD results

We present in this technical note an open-source web framework for the generation and visualization of parametric OpenFOAM simulations from surrogate models. It consists of a JavaScript module (rom.js) and a web app (cfd.xyz) to explore fluid dynamics problems efficiently and easily for a wide range of parameters. rom.js is a JavaScript port of a set of open-source packages (Eigen, Splinter, VTK/C++ and ITHACA-FV) to solve the online stage of reduced-order models (ROM) generated by the ITHACA-FV tool. It can be executed outside a web browser within a backend JavaScript runtime environment, or in a given web solution. This methodology can also be extended to methods using machine learning. The rom.js module was used in cfd.xyz, an open-source web service to deliver a collection of interactive CFD cases in a parametric space. The framework includes some tutorials, showing the whole process from the generation of the surrogate model to the web browser. It also includes a standalone web tool for visualizing users' ROMs by directly dragging and dropping the output folder of the offline stage. Beyond the current proof of technology, this enables a collaborative effort for the implementation of OpenFOAM surrogate models in applications demanding real-time solutions such as digital twins and other digital transformation technologies.

Sharing Species Pages through Living Atlases using Plinian Core: Current state and new developments

Living Atlases (living-atlases.gbif.org) is a community of developers and organizations built around the software platform developed by the Atlas of Living Australia (ALA, ala.org.au). ALA includes a number of open source reusable modules designed to assemble biodiversity data portals. Nowadays, more than 25 institutional and national biodiversity portals around the world have been implemented with the ALA tools (living-atlases.gbif.org/participants). The community organizes workshops and prepares training materials to present ALA modules to other institutions that want to implement it, to improve already existing portals and to learn from each other’s achievements (living-atlases.gbif.org/about). ALA based portals integrate data from multiple sources using biodiversity data standards ruled by TDWG, such as Darwin Core for species occurrences. The Living Atlases community is working in the integration of other data types, like species records and sampling events. Plinian Core (PliC) is a generalistic specification, oriented to share descriptions and nomenclature, as well as many other aspects (e.g. legal, conservation, and management details) of species level information from local and regional sources (Pando 2018). Following several years of development and earlier versions, PliC version 3.1 was made publicly available in 2012. The following year, after the approval of the “Species Information Interest Group'' by the TDWG Executive, a TDWG Task Group was created to start aligning Plinian Core with the TDWG Standards Documentation Standard (SDS). Several full-fledged implementations of Plinian Core were produced, between 2014 and 2019, including the Atlas of Living Costa Rica (CRBio.cr), the Cross-Nature Project (datos.iepnb.es), the Vasque Country (Spain) Official Nature Information System (ivap.euskadi.eus), the Colombian Biodiversity Catalog (catalogo.biodiversidad.co) and Enciclovida (Mexican Species Catalog, enciclovida.mx). These last three years, the Plinian Core Task Group has reviewed definitions of terms and formal declarations, assisted by the TDWG Technical Architecture Group (TAG), with the objective of ratifying PliC as a TDWG Standard. One of the most used tools to share standardized biodiversity data is the GBIF Integrated Publishing Toolkit (IPT, gbif.org/en/ipt). The IPT is an open source web application that allows institutions to standardize, to publish and share biodiversity data sets through Living Atlases, GBIF, and other biodiversity networks. The toolkit simplifies the process of publishing core data types (e.g. taxon occurrence, taxon checklists, sampling event) and, by means of extensions, it is possible to associate additional data with records of the core type. As part of our project, the extensions of PliC, accessible in the IPT Sandbox, were validated to show how they can be used by institutions to publish species pages through the IPT. In order to fully integrate Plinian Core based species pages into the ALA architecture, the Biodiversity Informatics Research Center (CRBio) has been working on the adoption of the Biodiversity Information Explorer (BIE), an ALA module which manages taxonomic and species contents by integrating global resources like EOL or Wikipedia. This adoption includes the required modifications of the Portal codebase to use the data model of the Plinian Core (github.com/tdwg/PlinianCore) to display a much richer content. This year, with assistance from the Capacity Enhancement Support Programme (CESP) from GBIF, we have revisited the integration of species pages information and expanded the content and type of information provided by the Atlas of Living Costa Rica. We present here the updated ALA Species Pages (ALA-BIE) and Index (BIE-Index) Modules and general documentation that could be applied to any ALA installation in order to prepare species records using the IPT and PliC standard, index and display them through the ALA-BIE Module.

Low-Cost, Open-Source, Emoncms-Based SCADA System for a Large Grid-Connected PV System.

This article describes a low-cost Supervisory Control and Data Acquisition (SCADA) system for a PV plant with local data logging. Typically, SCADA systems that are available on the market are proprietary (commercial), which are expensive and individually configured for a particular site. The main objective of this paper is to design a low-cost and open-source monitoring solution (hardware and software) to meet the requirements. The hardware used for this SCADA consisted of Arduino, Raspberry Pi, sensors, serial communication cables, and an open-source web view platform. This open-source platform manipulates, logs, and visualizes PV and environmental data. Emoncms runs on the Debian operating system. Field instruments were connected to two remote terminal units (RTUs). A PV array provided data to the RTU1, while an inverter output provided data to the RTU2, and the Raspberry Pi received the collected data in JSON format. As these data arrived, Emoncms used Emonhub as its main module, which refines data and then displays it on Emoncms’s WebView. The Raspberry Pi also stores data locally. Data logging was tested for 6 h, but the final results showed that data logging can last much longer. From an hour to a year, the data trend can be viewed on a user-friendly dashboard.

Workplaces: Pasts and Presents Takes the Digital Turn

<i>Workplaces: Pasts and Presents</i> Takes the Digital Turn

A web server for identifying circRNA-RBP variable-length binding sites based on stacked generalization ensemble deep learning network.

Circular RNA (circRNA) can exert biological functions by interacting with RNA-binding protein (RBP), and some deep learning-based methods have been developed to predict RBP binding sites on circRNA. However, most of these methods identify circRNA-RBP binding sites are only based on single data resource and cannot provide exact binding sites, only providing the probability value of a sequence fragment. To solve these problems, we propose a binding sites localization algorithm that fuses binding sites from multiple databases, and further design a stacked generalization ensemble deep learning model named CirRBP to identify RBP binding sites on circRNA. The CirRBP is trained by combining the binding sites from multiple databases and makes predictions by weighted aggregating the predictions of each sub-model. The results show that the CirRBP outperforms any sub-model and existing online prediction model. For better access to our research results, we develop an open-source web application called CRWS (CircRNA-RBP Web Server). Its back-end learning model of the CRWS is a stacked generalization ensemble learning model CirRBP based on different deep learning frameworks. Given a full-length circRNA or fragment sequence and a target RBP, the CRWS can analyze and provide the exact potential binding sites of the target RBP on the given sequence through the binding sites localization algorithm, and visualize it. In addition, the CRWS can discover the most widely distributed motif in each RBP dataset. Up to now, CRWS is the first significant online tool that uses multi-source data to train models and predict exact binding sites. CRWS is now publicly and freely available without login requirement at: http://www.bioinformatics.team.