Zhili Zhao

Reaction RuleML 1.0: standardized semantic reaction rules

RuleML is a family of XML languages whose modular system of schemas permits high-precision (Web) rule interchange. The familys top-level distinction is deliberation rules vs. reaction rules. In this paper we address the Reaction RuleML subfamily of RuleML and survey related work. Reaction RuleML is a standardized rule markup/serialization language and semantic interchange format for reaction rules and rule-based event processing. Reaction rules include distributed Complex Event Processing (CEP), Knowledge Representation (KR) calculi, as well as Event-Condition-Action (ECA) rules, Production (CA) rules, and Trigger (EA) rules. Reaction RuleML 1.0 incorporates this reactive spectrum of rules into RuleML employing a system of step-wise extensions of the Deliberation RuleML 1.0 foundation.

Hadoop MapReduce Framework to Implement Molecular Docking of Large-Scale Virtual Screening

Traditional virtual screening in the grid needs chemists to upload small molecule files and collect the results manually, which cannot implement docking and collection of results automatically. This caused heavy workload to chemists. In this paper, we took advantage of Hadoop platform in the massive data storage. We stored and managed small molecule files and docking results files using HDFS. In addition, MapReduce programming framework is used for parallel molecular docking to preliminarily process results files, in order to achieve the automation of the virtual screening molecular docking. The research of this thesis will be helpful to drug researcher by offering a massive data storage management system for large-scale virtual screening, and will also provide a reference for drug discovery in the cloud environment to promote the development of computational chemistry e-science.

Extending BPEL2.0 for Grid-Based Scientific Workflow Systems

SWF, short for scientific workflow, has recently emerged as a paradigm for orchestrating large-scale e-Science applications. SWF specification connects workflow designer with workflow engine, which makes it act ass one of the key components in SWF systems. We adopted BPEL as the Gird-based SWF specification because of its potential benefits to promote SWF sharing and reproducibility. However, grid-based SWF has unique features, which pure BPEL is not capable to handle. Appropriate ways to extend BPEL to fulfill the special needs should be found. We concluded three most necessary requirements and three kinds of alternative methods to extend BPEL within China Grid Support Platform, and implemented the method of adding additional abstractions upon BPEL to enhance user experience. Our work and findings suggest that our extension approach is feasible and would manifest potential advantage to SWF systems and finally innovate in the e-Science research and applications.

Principles of the SymposiumPlanner instantiations of rule responder

The Rule Responder SymposiumPlanner system supports topic-oriented collaboration between the distributed members of a virtual organization. Each member (or small team of members) is assisted by a semi-autonomous rule-based personal agent, which uses Semantic Web rules to capture aspects of the members (or teams) derivation and reaction logic. SymposiumPlanner is a series of Rule Responder use cases for supporting the RuleML Symposia (2007-2011) by coordinating personal agents that assist the symposium chairs, intelligently answering questions from people interested in the symposium. In this paper, we introduce principles of SymposiumPlanner and make suggestions about its future development, mainly for RuleML-2012, and about further Rule Responder use cases.

Using Deep Learning for Compound Selectivity Prediction

Compound selectivity prediction plays an important role in identifying potential compounds that bind to the target of interest with high affinity. However, there is still short of efficient and accurate computational approaches to analyze and predict compound selectivity. In this paper, we propose two methods to improve the compound selectivity prediction. We employ an improved multitask learning method in Neural Networks (NNs), which not only incorporates both activity and selectivity for other targets, but also uses a probabilistic classifier with a logistic regression. We further improve the compound selectivity prediction by using the multitask learning method in Deep Belief Networks (DBNs) which can build a distributed representation model and improve the generalization of the shared tasks. In addition, we assign different weights to the auxiliary tasks that are related to the primary selectivity prediction task. In contrast to other related work, our methods greatly improve the accuracy of the compound selectivity prediction, in particular, using the multitask learning in DBNs with modified weights obtains the best performance.

Event-Driven Scientific Workflow Execution

Scientific workflows streamline large-scale, complex scientific processes and enable different parts of a process to be systematically and efficiently executed on distributed resources. In this paper, we propose an event-driven framework for scientific workflows, which goes beyond the typical paradigm of global ECA (Event-Condition-Action) rules and executes scientific processes in terms of event message-driven conversations between rule agents. The behavioral reaction logic implemented by messaging reaction rules in combination with derivation rules used to represent complicated scientific conditional logic provides a highly expressive, scalable and flexible way to define complex scientific workflow patterns. Finally, a prototype system based on a Web rule engine Prova and a tool for rule-based collaboration Rule Responder is demonstrated.

Application-Objected Workflow Management System Based on Abstract Service

Since numerous large-scale scientific applications executed on Grids are expressed as complex scientific workflows, workflow management has appeared as one of the most important service in Grid environment in past few years. However, most of the management systems do not fully hide the complexity of the underlying orchestration language. Meanwhile, Grid environment is dynamic when each scientist involves in different requirements to submit their computational jobs to heterogeneous executing environment. Therefore, in order to solve the issues of complexity, heterogeneity and dynamism, in this paper we developed Application- objected Workflow Management System based on Abstract Service. The Abstract Service is an abstract combination service for a group web service with similar function. And we presented Service Center, which contains a large number of abstract services to form a virtual uniform view to shield the diversity among various web services. We also developed a visual workflow designer, which allows scientists to orchestrate an abstract workflow by just dragging or dropping the graphical representations of abstract services to (or from) the canvas. Especially, the Requirement Description Document (RDD) was proposed to describe the special demands of individual computational jobs in XML grammar. And we implemented a Workflow Application Management System (WFAMS), which translates abstract workflow in well-defined order to concrete workflow, and manages, controls and decomposes the workflow, and then combines the scientist’s relational data and submits workflow to job pool in Grid platform. Finally, the solutions to the key problems in the workflow management system are considered, including the definition of abstract workflow, the mechanism to support failure handling, the workflow template and the mechanism to isolate the control flow and data flow.

An Extensible Database Management System for Large-Scale Virtual Screening

Virtual Screening (VS), most well-known method which can improve the efficiency greatly in drug discovery, involving millions of small molecules in the process with one target enzyme. As a data intensive application, how to store and access those ever-increasing data for the corresponding Grid jobs running in parallel so as to improve the efficiency of VS has not been thoroughly explored. In this paper we proposed a data allocation strategy to cope with it. Besides, we implemented an extensible database management system for VS. The flexible data processing servers in data pocessing together with flexible DB servers in data storage are significant components for the system extensibility. At last, a comprehensive performance evaluation of the system is given.

A User-Friendly Task Editor Environment for Large-scale Virtual Screening Application

Virtual Screening (VS) is a large-scale parallel Bag-of-Tasks application whose tasks are independent with no data dependency and communication. Grid computing facilitates VS by providing shared computational resources across different international institutions. However, the traditional grid middle wares do not provide mechanism of task decomposition, and users usually need to decompose the large-scale independent tasks into small sub-tasks manually to make the most of the heterogeneous resources. Such method can be complex and may place a burden on the users. In this paper we propose a user-friendly Task Editor Environment extending JSDL for VS, including GUI Designer Data Description Module and Task Editor Module. The advantage of the proposal is to separate the data description from task definition description and ensure the task execution optimism. At last we present a case study, which turns out that our proposal not only satisfies the users’ need to perform the VS application in grid environment with a Task Editor Environment, but also improves the efficiency of VS application.

ChemDataBase 2: An Enhanced Chemical Database Management System for Virtual Screening

Virtual screening is a newly emerging method used in new drug research and development (R&D). The introduction of virtual screening immensely reduces R&D cycle and decreases R&D cost directly. Mass of chemical data used for virtual screening needs to be managed effectively in certain ways. We have developed the first release of ChemDataBase, but it only has the basic functions of chemical database which can not meet requirement of virtual screening fully. This paper investigates and implements the second and improved release of ChemDataBase which is better suitable for virtual screening. Compared with the first release, the second release has several new features. For one thing, it is based on CSGrid and supports accessing of Grid database created in CSGrid. For another thing, it adopts Hibernate framework for data persistence management. In addition, it can invoke data analysis software to analyze the results of virtual screening experiment. Finally, a case study in computational chemistry shows that the second release plays an important role in new drug R&D.