FIll in sections below, then cut and paste into the report on research.gov. Due date: September 30th, 2014. I use Heading 1, 2, etc for the report sections. I use [TEXT GOES HERE] to indicate where you should put text. Text in red is instructions from reports.gov. ​ Milestones: please see what we submitted to Dan K and address these in the report: ​ https://docs.google.com/document/d/1Z7tuKSUjZCCQpwQOSIq3oURTfnHnWGFWx8-uIP139A w/edit#heading=h.jch3hpwyme5

For outreach, see https://docs.google.com/document/d/1KsgYZqF3aLU6Vs0Xf1i2Gu9_k5wlW81hxBNgjCt3Fg8/ed it#heading=h.v0qfwhj3h1bw Accomplishments For NSF purposes, the PI should provide accomplishments in the context of the NSF merit review criteria of intellectual merit and broader impacts, and program specific review criteria specified in the solicitation. Please include any transformative outcomes or unanticipated discoveries as part of the Accomplishment section. What are the major goals of the project? 8000 char limit.

List the major goals of the project as stated in the approved application or as approved by the agency. If the application lists milestones/target dates for important activities or phases of the project, identify these dates and show actual completion dates or the percentage of completion.

The overall goal of the SciGaP project is to develop a Science Gateway Platform as a service that can be used to power both participating gateways (UltraScan, CIPRES, NSG) as well as gateway community at large. To meet this, goal, we are implementing SciGaP as a multi-tenanted service hosted on a virtualized environment that provides a well-defined Application Programming Interface (API) for client gateways. Gateways can use SciGaP through this API to outsource the generic capabilities such as task execution on supercomputers and workflow management.

Highlights this year include the XSEDE14 SciGaP tutorial, which provided focus for our development efforts and a highly visable venue for interacting with the science gateway community; the design and implementation of the Airavata API based on the analysis of a broad array of science gateway use cases compiled for XSEDE (https://www.ideals.illinois.edu/handle/2142/43883) and development patterns (Marru, S.; ​ ​ Dooley, R.; Wilkins-Diehr, N.; Pierce, M.; Miller, M.; Pamidighantam, S.; Wernert, J., "Authoring a Science Gateway Cookbook," Cluster Computing (CLUSTER), 2013 IEEE International Conference on , vol., no., pp.1,3, 23-27 Sept. 2013)); the development of a reference client implementation in PHP to provide a face for the API and an entry point for new developers; the deployment of the initial set of SciGaP services to support the XSEDE14 tutorial; and the integration of Airavata with the UltraScan Science Gateway.

The following are the project’s recurring and Year 1 milestones. We summarize activities related to milestones as sub-bullets. Recurring Deliverables

● Engage in planned outreach activities ○ Presented Airavata API design and community architecture engagement at ApacheCon 2014, these API’s have resulted from SciGaP interactions. ○ Conducted gateway developer tutorial at CCGrid2014. ○ Conducted gateway developer tutorial at XSEDE14. ○ Pierce and Marru are participating in two Fall 2014 semester classes taught by Prof. Geoffrey Fox and Prof. Beth Plale, respectively. Course materials will be made available online as part of the IU School of Informatics and Computing’s Online Course strategy. ○ We mentored five Google Summer of Code 2014 student projects, including one with participation by PI Demeler (in addition to IU team). Details are listed here: https://cwiki.apache.org/confluence/display/AIRAVATA/GSoC+2014 ○ We are engaging new students in IU’s School of Informatics and Computing through “bootcamp” activities. Over 25 students have participated in overview presentations and are subscribed to the [email protected] mailing list. ​ ​ ○ The CIPRES project engaged one student from a women-only high school, who matriculated to UCLA in molecular biology in 9/2014. The CIPRES project also provided internships for two undergraduate students and two Master’s level students, all in computer science. One of these students has entered a PhD program at UIUC in August, 2014. ○ Three high school students participated in summer internship and worked on various NSG projects as a part of SDSC’s Research Experience for High School (REHS) program. NSG team hosted workshops at Society for Neuroscience (SFN November , 2013) annual meeting, at NEURON workshop ( June, 2014), and at Computational Neuroscience (CNS July, 2014) annual meeting. SFN and CNS had about 22 to 25 attendees and the NEURON workshop had 12 attendees. In addition NSG poster was accepted at both of these conferences and were presented by NSG team members. ● Conduct yearly gateway community user surveys. ○ The CIPRES project conducted a user survey in March to assess progress and trends in user needs. This year survey was emailed to 3200 users, and 800+ responses were received. ○ IU team members helped design and implement a major gateway survey as part of the related Science Gateway Institute, an NSF S2I2 project led by Nancy Wilkins-Diehr. An initial summary report was submitted to the NSF by the Institute. ● Follow 6-week release pattern for SciGaP software patches. ○ We released Apache Airavata 0.8-0.13 during the current period, with release 0.14 in preparation (anticipated by end of September).

Year 1: SciGaP 1.0 Foundations ● Release Airavata 1.0 with API based on design documents and community input. ○ The Airavata API has been defined and validated through two tutorials, four Google Summer of Code projects, and integration with UltraScan. We are preparing release 1.0 based on this work and anticipate release in late October 2014, pending community vote. ● Design and begin implementing multi-tenanted SciGaP. ○ We have one accepted publication to CCGrid 2014 describing our approach to multi-tenanted security credential management, one accepted publication to the International Workshop on Science Gateways (2014) on the overall architecture, and one submitted publication to Gateway Computing Environments Workshop describing SciGaP’s multi-tenanted, API based approach and its implementation to support UltraScan’s usage requirements. ● Convert UltraScan to use SciGaP. ○ We completed conversion of UltraScan from legacy, pre-Airavata code for job submission and management in December 2013 but maintained the pre-existing interfaces. We are in the process of converting UltraScan-Airavata integration to use the new Airavata API and expect production usage by the end of September 2014, coinciding with a tutorial by PI Borries Demeler. ● Begin conversion of NSG to SciGaP. ○ NSG and CIPRES are branches of the same code base (the Workbench Framework). During Year 1, we have developed a common code base. CIPRES is already using this code base and NSG is in the process of adopting this code base. From analysis of this common code base, we derived two possible strategies for integrating the Workbench with Airavata. This was presented as a poster at XSEDE14. We also determined that it would be better to focus on integrating Airavata with this common code base, with evaluation through the CIPRES gateway, instead of NSG integration in Year 2. ● Initiate Net+ evaluation (see http://www.internet2.edu/netplus/service-phases.html). ​ ​ ○ We met with the Net+ team via teleconference and presented our project. It was determined that the Net+ team did not have the resources to help SciGaP develop the necessary business plan. Instead, the Net+ team recommended that we consult with Dr. Brad Wheeler (Indiana University’s Vice President for Information Technology and co-founder of Internet2’s Net+ activity) for recommendations and possible collaborations with faculty in Indiana University’s Kelley School of Business. We will follow up on this recommendation to meet with Dr. Wheeler.

What was accomplished under these goals (you must provide information for at least one of the 4 categories below)? 8000 char limit for each of the 4 categories. For this reporting period describe: 1) major activities; 2) specific objectives; 3) significant results, including major findings, developments, or conclusions (both positive and negative); and 4) key outcomes or other achievements. Include a discussion of stated goals not met.

1) Major activities

Project Wide: Our project-wide focuses this year were on developing a SciGaP API based on ​ detailed understanding of the CIPRES, NSG, and UltraScan use cases; implementing the API in Apache Airavata; standing up a preliminary version of the persistent, unified SciGaP services, and integrating UltraScan with these persistent services. In order to bootstrap the new collaboration between IU and SDSC team members, IU staff programmer Don Kushan Saminda Wijeratne was embedded with the SDSC development team. This effort is summarized in the XSEDE14 poster, “Evaluating CIPRES Science Gateway Integration Options with Apache Airavata” by Don Kushan Wijeratne, Mark Miller, Terri Schwartz, Suresh Marru and Marlon Pierce.

IU and UltraScan team members also collaborated extensively on the API design and initial implementation in Apache Thrift with PHP client bindings. A two-day hackathon with UltraScan’s Gary Gorbet and the IU team in Bloomington in March 2014 was the keystone of this effort. Final integration with the API is being completed at the end of September 2014, coinciding with a tutorial in Juelich by UltraScan PI Borries Demeler.

Apache Airavata: We released Airavata 0.10-0.13 through the Apache Software Foundation, ​ with 0.14 in preparation for the end of the month. Our major software development activities focused on defining the Apache Airavata API using Apache Thrift; a complete redesign of the Airavata Registry component to replace legacy data model definitions with data model bindings defined by Thrift; adding a new major component, the Orchestrator, to handle scheduling and simplify support for UltraScan, CIPRES, and NSG use cases; rewiring Airavata components to use the API Server and the Orchestrator; and reimplementing Airavata’s GFAC component to make it more fault-tolerant and elastic through the use of Apache Zookeeper. This work is tracked through Apache Airavata mailing lists and and Airavata’s Jira issue tracking system.

In addition to Airavata work, we created a GitHub area, https://github.com/SciGaP/, for SciGaP ​ ​ and related activities that did not fit specifically in Airavata. The major component here is the PHP Reference Gateway, which we developed to illustrate the Airavata API’s features during tutorials and outreach activities, validate the API’s usage for PHP clients, and potentially serve as the basis for future science gateways.

In addition to this development work, we engaged the Apache, XSEDE, and international Science Gateway communities. We organized two peer-reviewed science gateway tutorials (CCGrid 2014 and XSEDE 2014). We presented Airavata and SciGaP work at the following community conferences. Associated papers are listed under publications.

● ApacheCon 2014: “RESTless API Design with Apache Thrift: Experiences from ​ Rearchitecting Apache Airavata”. This presentation described our work designing the ​ Airavata API. http://sched.co/MnP7VB ​ ● ApacheCon 2014: “Apache Airavata: Building an Elastic Platform as a Service Using the ​ Apache Ecosystem”. http://sched.co/1bsTVow ​ ​ ● CCGrid 2014: “A Credential Store for Multi-Tenant Science Gateways”, presentation of peer-reviewed conference paper. This work was partially supported by the SciGaP award and partially by the NSF award “SDCI Sec: Distributed Web Security for Science Gateways” (Jim Basney, PI). ● International Workshop on Science Gateways 2014: “Apache Airavata: Design and Directions of a Science Gateway Framework”, presentation of peer-reviewed paper. https://sites.google.com/a/my.westminster.ac.uk/iwsg2014/home/dates/apache_airavata ● XSEDE14: “Evaluating CIPRES Science Gateway Integration Options with Apache Airavata”, Presentation of peer-reviewed poster. https://www.xsede.org/documents/527334/743323/xsede14_submission_253.pdf

The IU team is also engaging in on-campus outreach to graduate students in the School of Informatics and Computing. We are giving lectures and sponsoring projects in Prof. Beth Plale’s and Prof. Geoffrey Fox’s “Topics in Informatics” courses, and we are in discussions with Prof. Judy Qiu for participating in her course during the Fall 2014 semester. We also offered an informal “Airavata Bootcamp” course on September 1, 2014 that was attended by approximately 25 students; 28 students have signed up to the [email protected] mailing list. ​ ​ Airavata’s use of Apache Thrift as a means to expose its public programming interfaces and associated with data models is novel in the science gateway community and so requires careful review. Initial prototyping of roles and role management through the API using the WSO2 Identity Server was the subject of a GSOC 2014 student project (http://www.google-melange.com/gsoc/proposal/public/google/gsoc2014/scnakandala/57273904 ​ 28823552). We are also collaborating closely with the Center for Trustworthy Scientific ​ Cyberinfrastructure (CTSC) (Von Welch, PI and Randy Heiland, lead consultant). This has resulted in an initial report and set of recommendations; in particular, we will be evaluating the use of Thrift over HTTPS instead of the TCP/IP bindings it normally uses in order to leverage extensive pre-existing security. We have also adopted a mutually agreed upon continuing engagement arrangement with CTSC. Our next target is understanding the security requirements for the three-way interactions between client gateways, SciGaP’s multi-tenanted Airavata services, and SciGaP’s Identity Server.

UltraScan Gateway: Major activities in the UltraScan gateway occurred in multiple areas. ​ These included 1) the development of new analysis methods for analytical ultracentrifugation (AUC) users, and their implementation in the desktop GUI code base; 2) efforts to improve gateway accessibility in Europe; 3) improvements in the Airavata integration logic; 4) improving and extending LIMS functionality; 5) in the support of new data types obtained from recently developed instrument detectors and novel experiment types; and 6) in the gateway expansion to new user communities.

CIPRES Science Gateway: Major activities in the CIPRES Science Gateway included 1) the ​ creation of a distributable software package that can be used to create a new Gateway, 2) planning how this package will be integrated with Airavata over time, and 3) working together with the Airavata developers to plan the integration, and providing driving requirements for the Airavata API.

Neuroscience Gateway: Major activities of the Neuroscience Gateway (NSG) include adding ​ new features as requested by the user community, providing feedback regarding specific NSG requirements for the SciGaP project, outreach to the user community, and involving high school students for internship. Examples of new features include 1) adding parameter sweep option such that users can bundle up multiple jobs each with different input parameter, 2) providing command line option, via the gateway portal interface, for one of the neuronal simulation software. Specific requirements of the NSG were provided for the SciGaP project; example of this include the requirement to compile codes describing network models and then use that with neuronal software package for time dependent simulations. Multiple workshops, and poster presentations were done as a part of the outreach, and three high school students participated as student interns during the summer of 2014.

2) Specific Objectives Our project’s specific objectives are described under “Major Goals”. In summary, our objectives this year were to define and validate a single API for all Airavata components, integrate it with the UltraScan gateway, and investigate strategies for integrating with the Workbench Framework-based CIPRES and Neuroscience Gateways.

3) Significant results, including major findings, developments or conclusions Project Wide: Our most significant results were the transition of UltraScan to use Airavata ​ services (replacing legacy, pre-Airavata software versions), the definition and implementation of the Airavata API based on UltraScan, CIPRES, and NSG requirements, and the deployment of an initial set of unified SciGaP services on IU’s Quarry Gateway Hosting System to support both UltraScan and tutorial usage.

Apache Airavata: Our primary results were a) the demonstration that Apache Thrift could be ​ used to develop public Application Programming Interfaces with client software development kits in Java, PHP, and C++; b) the feasibility of defining and implementing our data models using Apache Thrift; c) initial determination that we can use Apache Zookeeper for internal state management of tasks managed by the GFAC component, allowing us to provide failover services; d) that these significant changes to Apache Airavata are production ready through integration with the UltraScan Science Gateway. We also began preliminary evaluations of the ways to integrate Airavata with CIPRES and NSG; this was closely linked to the realignment of the CIPRES and NSG code bases.

UltraScan Gateway: One of our findings was related to gateway performance. We discovered ​ that long latencies observed abroad played a major role in the user experience of our gateway. Users accessing our gateway from Europe for example experienced up to 20 fold slower interactions with the gateway database. We traced these issues to platform-specific performance issues (In particular, Windows failed to properly use transmission protocols used on transatlantic Internet connections), and general slowness experienced when performing MySQL searches across the Atlantic. Our solution to restore gateway performance was to establish a collaboration with Juelich Supercomputing Center in Germany and colocate a virtual machine functioning as the gateway LIMS server for users in Europe. By making this change we have restored user experience to speeds comparable to those observed in the US. We plan to implement additional gateway servers also in India and other continents to provide a seamless user experience worldwide. The implications of these requirements on the SciGaP API and implementation are discussed in our submission to the Gateway Computing Environments 2014 Workshop (currently under review).

Analysis Performance Prediction: We have developed very accurate prediction models that will allow us to predict the CPU time a particular data set and analysis parameter set will require for completion on XSEDE resources. These prediction models vary for Trestles, Stampede and Lonestar. We plan to implement the prediction algorithms to enhance queuing times and allow gateway operators to obtain more efficient scheduling performance for UltraScan Gateway jobs.

We have developed two new major analysis methods in the past year that significantly enhance the analysis capabilities and range of applicability within UltraScan. One of the methods, the Custom Grid method, targets a new area of research for AUC, material science, nanoparticles and synthetic polymers to solve long-standing problems with density prediction. Gateway methods were developed to support this new analysis approach on the XSEDE infrastructure.

A major development relates to the incorporation of new datatypes generated by a novel multiwavelength detector for the analytical ultracentrifuge. This detector adds a spectral dimension to the experimental data, opening avenues for enhanced multi-dimensional analysis of AUC data. This data format also significantly increases the data density and information content by about 300-500 fold. This required major adjustments in the gateway interface and the supporting desktop software. At this point, we have the basic infrastructure in place to analyze multiwavelength data on the UltraScan Gateway, and are able to support all UltraScan data analysis methods performed on supercomputers with the gateway functionality. Further improvements in job meta-scheduling are desirable to allow optimized workflow designs and will be tackled in the coming year. This requirement makes a general purpose job throttling and scheduling capability a very high priority for SciGaP services. Implementation is underway in Apache Airavata.

NeuroScience and CIPRES Science Gateway: Unifying the code bases of CIPRES and the ​ Neuroscience Gateways was another major project effort completed this year. This unification will simplify the process for both gateways, as well as other gateways derived from this common framework, to use SciGaP services. The CIPRES and Neuroscience Gateways are built from a ​ ​ gateway software package that consists of two parts: the Workbench Framework (WF, which manages the executive part of Gateway functioning) and the Web Application (WA, which creates the browser GUI for the Gateways). The PoPLAR Science Gateway is also based upon this software package. While all three Gateways were created from the same software package, they were created independently, and each contain a significant number of hard-coded references to project-specific features. After some deliberation, we decided that the best course of action for integrating these two Gateways with the Airavata software framework was to first refactor the WA/WF software packages to create a single, distributable open-source software package, and then devise a strategy to integrate this software package with the Airavata software.

The decision was driven by the fact that both the CIPRES and NSG projects need to be integrated with Airavata, but both projects are operating in a full production mode, and are under rapid development, for e.g. adding REST access under support from other NSF funding for the CIPRES project, and incorporating new features of parameter sweep option, based on user requirements, for the NSG project. Integrating these Gateways under a single distributable package will make it easier for the NSG and CIPRES Gateways to integrate with the Airavata framework, without disrupting current forward progress. It will also make it easier for other Gateways that are based on CIPRES can also take advantage of integration with AIravata and the newly developed REST services simply by adopting the distributable WF/WA software package.

The refactoring process has now been completed. This means all references specific to a particular Gateway (for example, location of the project database, bug tracker, and text messages) are now specified by a set of configuration files that are controlled by the person(s) installing the WA/WF software locally. This package is currently supporting the test and production versions of the CIPRES Science Gateway and the PoPLAR Science Gateway, and work is ongoing for the NSG project to use package also.

We are now working actively with the Airavata development group to integrate the WA/WF packages with the Airavata software. We will manage the transition to use of Airavata slowly, adopting first the Job submission and monitoring features of Airavata. As the the Airavata job submission software is load tested, we will but this feature into production for CIPRES and the NSG. We will continue to adopt other features available through Airavata as the API of the Airavata package becomes stable. The strategy we have followed will allow users to adopt Airavata as part of their configuration if they choose to use the WA/WF software to create their gateways.

4) Key outcomes or other achievements. Our key outcome is the definition and implementation of the Airavata API and data models based on a thorough examination of the UltraScan, NSG, and CIPRES use cases. In addition, we have developed and released a simple prototype science gateway to illustrate the use of the API and serve as a starting point for both student projects and new gateway spinoffs.

We have also evaluated using the API to integrate SciGaP with the work of Dr. Emre Brookes, co-investigator on CCP-SAS project (ccpsas.org), an NSF/EPSRC (US/UK) funded collaboration. Brookes, a former student of SciGaP PI Demeler, is building GenApp, a tool for rapidly building user interfaces to a variety of complementary tools for molecular structure determination. Integration of GenApp with Airavata was the subject of a GSOC 2014 project (https://cwiki.apache.org/confluence/display/AIRAVATA/GSoC+2014+-+GenApp+Integration+wi ​ th+Apache+Airavata) and resulted in a paper submission to the Gateway Computing ​ Environments Workshop, currently under review.

SciGaP team members have also been involved in service to the science gateway community, serving as co-organizers of the Science Gateways Workshop at Cluster 2013 and the Gateway Computing Environments 2014 workshop at Supercomputing 2014. SciGaP team members have also served on review panels for the International Science Gateway Workshop (IWSG) 2014 and XSEDE 2014. What opportunities for training and professional development has the project provided? 8000 char limit.

Describe opportunities for training and professional development provided to anyone who worked on the project or anyone who was involved in the activities supported by the project. "Training" activities are those in which individuals with ​ advanced professional skills and experience assist others in attaining greater proficiency.

If the research is not intended to provide training and professional development opportunities or there is nothing significant to report during this reporting period, please check "Nothing to Report" if applicable.

Project Wide: We organized a SciGaP tutorial for XSEDE 2014. The XSEDE 2014 tutorial was ​ attended by approximately 15-20 participants. Marlon Pierce, Suresh Marru, David Reagan, and Mark Miller were the presenters, and the UltraScan gateway was used to demonstrate a science gateway.

As a member of the Apache Software Foundation, Apache Airavata can chose from the allotment of Google Summer of Code positions given to the Apache Software Foundation. This greatly increases our ability to get students. In GSOC 2014, we had one student working on a joint UltraScan-Airavata project for job scheduling (https://cwiki.apache.org/confluence/display/AIRAVATA/GSoC+2014+-+Prototype+Airavata+Su ​ pport+for+Application+Scheduling). We will continue this in 2015, identifying GSOC students ​ who can contribute to multiple SciGaP partners.

Apache Airavata: We organized a tutorial on Apache Airavata at CCGrid 2014 (Chicago, May ​ 2014) as a precursor to the XSEDE 2014 SciGaP tutorial.

UltraScan Gateway: We have presented multiple workshops where we trained gateway user ​ communities in the use of our software and helped them learn to use the Gateway for their research objectives. Workshops that were presented during year one included:

1. UltraScan workshop at the 21st international Analytical Ultracentrifugation Conference in Atami, Japan. This workshop was attended by approximately 20 users from the US, Europe (Switzerland, Germany, Great Britain), Australia, Japan, Taiwan and China. 2. UltraScan workshop at the EMBO Practical Course: Modern Biophysical Methods for Protein-Ligand Interactions at Oulu, Finland. This course was offered to 20 international students from Germany, Israel, Italy, Estonia, Slovenia, Czech Republic, Finland, Portugal, Norway, Switzerland, Brazil, South Africa, Russia, France and Spain. 3. UltraScan workshop at the Max Planck Institute for Biophysics in Munich, Germany for 12 students from Germany and Switzerland. 4. 12th Annual UltraScan workshop at the University of Texas Health Science Center at San Antonio to 10 graduate students, postdoctoral candidates, technicians and summer intern students from the UTHSCSA Biochemistry department. The students came from various institutions in the US, including Natl. Institute of Standards, Colorado State University, Florida State Univ., North Carolina State University, University of Michigan, and one student from KAUST, Saudi Arabia. (http://www.ultrascan.uthscsa.edu/workshop/) ​ ​ In addition, we trained two high school students as part of the eight-week long San Antonio Northside Independent School District’s Summer Internship Program and three students from the three-year long UTHSCSA Voelcker Biomedical Research Academy program. One of the NISD students (Haram Kim) developed a performance prediction software program by data mining of performance data stored in the UltraScan Gateway database for over 70 LIMS instances. The results will be used to improve scheduling performance on three XSEDE clusters (Lonestar, Stampede, and Trestles) of the 2-dimensional spectrum analysis method which is heavily used by all Gateway users. All other students were taught AUC data analysis with UltraScan and trained in biophysical laboratory skills, and gathered data for a new experimental protocol developed in our laboratory.

CIPRES Science Gateway: Nothing to report at this time. ​ Neuroscience Gateway: We have presented the NSG at multiple workshops to train users ​ about how to use the gateway and also to collect feedback from users regarding features that users wanted for the future.

1) NSG workshop at the Society for Neuroscience (SFN) annual meeting In November 2013 in San Diego. This NSG workshop was a half day workshop and included presentation and hands-on session on NSG as well as talks by some of the NSG users. About 22 attendees attended the workshop.

2) We presented a talk on NSG at the NEURON 2014 summer course at the University of California San Diego in June 2013. About 12 attendees were present at this summer course.

3) We hosted a NSG workshop at the Computational Neuroscience Society 2014 conference in Quebec City, Canada. About 25 attendees attended this workshop and it included presentation and hands-on demo on NSG, a talk by a developer of a neuronal simulation code (Brian) provided by NSG, and two user talks.

As a part of SDSC’s Research Experience for High School (REHS) students program, three high school students did internship at SDSC during the summer of 2014. They worked on various projects related to the NSG. How have the results been disseminated to communities of interest? 8000 char limit.

Describe how the results have been disseminated to communities of interest. Include any outreach activities that have been undertaken to reach members of communities who are not usually aware of these research activities, for the purpose of enhancing public understanding and increasing interest in learning and careers in science, technology, and the humanities.

Project Wide: We engaged the community through the XSEDE14 tutorial, which covered ​ science gateways in general, Apache Airavata, UltraScan and CIPRES. We have one peer-reviewed poster in collaboration with CIPRES and NSG team members. We have one paper co-authored by Airavata and UltraScan team members submitted to GCE14 that is under review. If accepted, this will be presented at the SC14 workshop.

The assistance we have received from many collaborators is complementary to our dissemination efforts and is also described below. This two-way engagement with scientific communities, software developers, and resource providers is essential for project success.

Apache Airavata: We have two accepted conference papers on Apache Airavata’s architecture ​ as it moves toward SciGaP’s goals. The Airavata team members give frequent presentations to the community. These include talks at the CCP-SAS Project Kickoff Meeting (February 2014), two presentations at ApacheCon 2014 (March 2014); and a colloquium at the IUPUI School of Computer Science (September 2014).

We also identified the need within the Apache Airavata project to engage with the broader Apache and cyberinfrastructure communities outside the usual “developer” mailing list. We created an “architecture” mailing list in order to promote more higher level discussion about Airavata’s design and goals. We featured the architecture mailing list and some of its benefits at one of our ApacheCon 2014 presentations.

UltraScan Gateway: Our dissemination activities encompass multiple approaches. First, we ​ distribute our software from our website (http://www.ultrascan.uthscsa.edu) where AUC ​ ​ practitioners tend to first land to look for data analysis software for their AUC experiments. This software is available free of charge, pre-compiled for Linux, Windows and Macintosh. Furthermore, we offer new Gateway instances through a web form, where users at previously not served research institutions can request a new LIMS instance (http://uslims3.uthscsa.edu/uslims3_newlims/request_new_instance.php) or sign up on an ​ ​ existing gateway instance (http://uslims3.uthscsa.edu/uslims3_newlims/newaccount.php). ​ ​ Furthermore, we offer workshops in the US or abroad, and in conjunction with national and international conferences where we train new gateway users and advertise the gateway. Furthermore, the PI directs the Center for Analytical Ultracentrifugation of Macromolecular Assemblies (CAUMA, http://www.cauma.uthscsa.edu/) which is the premier service laboratory ​ ​ for AUC experiments in the US, making this technology available to investigators from across the country. Many of these users learn about our software and use the gateway through data sharing and ultimately become new users.

Our biophysical training efforts (described above) further help to disseminate our software and attract new users.

The UltraScan Gateway was helped tremendously from in-person support received from various users who provided data, testing results, and feedback, and developers who made significant UltraScan code contributions and improvements in Airavata through the Unicore integration. These individuals include Shabhaz Memon, Morris Riedel, Norbert Attig, Thomas Lippert, Florian Janetzko (Juelich Supercomputing Center), Konrad Loehr, Johannes Walter (University of Erlangen, Germany), Dirk Haffke, Helmut Coelfen (University of Konstanz, Germany), Luitgard Nagel-Steger, Martin Wolff (University of Duesseldorf, Germany)

CIPRES Science Gateway: The distributable software package the supports the CIPRES ​ Science Gateway is freely available to other Gateway projects, currently through the read only SVN repository at SDSC. http://svn.sdsc.edu/repo/NGBW/cipres-portal/ ​ The CIPRES Science Gateway has collaborated with a number of software developers to improve, configure, and release new codes into production. These individuals include Alexandrous Stamatakis (Heidelberg Institute for Theoretical Studies), Tracy Heath (University of California, Berkeley), Nicolas Lartillot (University of Montreal), Diego Darriba (Universidade de Vigo), Andrew Meade (University of Reading), Rob Lanfear (Macquarie University, Sydney) and Remco Bouckaert (University of Auckland). In addition, we have worked with 12 individual scientists to make custom runs outside the CIPRES interface. This activity helps us better understand where the deficiencies in our interface lie, and the requirements for the next wave of phylogenetics scientists. Finally, we have worked with dozens of users over the past year to improve our tool interfaces, and better meet individual needs.

Neuroscience Gateway: As mentioned earlier the NSG is based on the Workbench Framework ​ (WF) and Web Application (WA) and required incorporating some new features specific for the neuroscience community. The software package is available freely to anyone interested.

We have mentioned the three workshops we hosted in 2013 and 2014. In addition presentations are made at Extreme Science and Engineering Discovery Environment (XSEDE) and International Workshop on Science Gateways (IWSG) annual conferences and through special journal issue publications such as in the May 2014 special issue publication on science gateways in the journal of Concurrency and Computation Practice and Experience.

Developers of neuronal simulations tools have provided valuable help and feedback to the NSG project staff. This includes Michael Hines (Yale University), Dave Beeman (University of Colorado), Padraig Gleeson (University College London), Upi Bhalla (National Center for Biological Sciences, TIFR, Bangalore, India), Marcel Stimberg (Institut de la Vision, Paris, France), Dan Goodman (Imperial College London), Andrew Davison (Centre National de la Recherche Scientifique, Paris), and Markus Diesmann (Theoretical Neuroscience, Julich Research Centre, Germany).

What do you plan to do during the next reporting period to accomplish the goals? 8000 char limit Supporting files. You may upload pdf files with images, tables, charts, or other graphics in support of this section. You may upload up to 4 pdf files with a maximum file size of 5 MB each.

Our project milestones remain unchanged from the original milestones and outreach plans submitted at the start of the project, with one modification: we will focus on integrating the common Workbench Framework code base (that aligned NSG and CIPRES) and Airavata and testing with CIPRES in Year 2 rather than on NSG-Airavata integration. Since both CIPRES and NSG depend on the newly unified Workbench Framework common code base, this does not represent a major change

We additionally plan to finalize the Airavata 1.0 release (a carry-forward of a Year 1 goal) and to finalize the API’s security model in consultation with CTSC. The security model was not previously listed as a milestone but will be a significant activity in Year 2 that will be required for multi-tenancy. Products

For NSF purposes, the PI should include and discuss in the Product section the goals associated with data management and access and note any significant changes in them, as well as specific plans for dissemination of data, software and other digital research products. When you report any of these items, please include any available identifiers and whether and how these products can be accessed or shared.

INSTRUCTIONS - List any products resulting from the project during the reporting period. ​ If there is nothing to report under a particular item, please check, "Nothing to Report" if applicable.

Your Output Summary for this Reporting Period: Publications (0 items) For NSF purposes, eac​ h category of publication should identify any associated data, software, other supplementary material and their appropriate identifiers. "Other publications, conference papers and presentations" should include other "non-reviewed" publications, conference papers, and presentations.

1. Swygert SG, Manning BJ, Senapati S, Kaur P, Lindsay S, Demeler B, Peterson CL. Solution-state conformation and stoichiometry of yeast Sir3 heterochromatin fibres. Nat ​ Commun. 2014 Aug 28;5:4751. doi: 10.1038/ncomms5751. 2. Demeler B., Nguyen TL, Gorbet GE, Schirf V, Brookes EH, Mulvaney P, El-Ballouli AO, Pan J, Bakr OM, Demeler AK, Hernandez Uribe BI, Bhattarai N, and RL Whetten. Characterization of Size, Anisotropy, and Density Heterogeneity of Nanoparticles by Sedimentation Velocity. Anal Chem. 2014 Aug 5;86(15):7688-95. ​ 3. Memon S., M. Riedel, F. Janetzko, B. Demeler, G. Gorbet, S. Marru, A. Grimshaw, L. Gunathilake, Raminder Singh, N. Attig and T. Lippert. Advancements of the UltraScan ​ ​ scientific gateway for open standards-based cyberinfrastructures. Concurrency ​ Computat.: Pract. Exper. (2014) Wiley. DOI: 10.1002/cpe.3251 4. Gorbet G., T. Devlin, B. Hernandez Uribe, A. K. Demeler, Z. Lindsey, S. Ganji, S. Breton, L. Weise-Cross, E.M. Lafer, E.H. Brookes, B. Demeler. A parametrically constrained ​ ​ optimization method for fitting sedimentation velocity experiments. Biophys. J. (2014) Vol ​ 106(8), pp1741-1750 dx.doi.org/10.1016/j.bpj.2014.02.022 5. Pham JD, Demeler B, Nowick JS. Polymorphism of Oligomers of a Peptide from ​ ​ β-Amyloid.. J Am Chem Soc. 2014 Mar 26. PMID: 24669785 ​ 6. Brandariz-Nuñez A, Valle-Casuso JC, White TE, Nguyen L, Bhattacharya A, Wang Z, Demeler B, Amie S, Knowlton C, Kim B, Ivanov DN, Diaz-Griffero F. Contribution of ​ ​ oligomerization to the anti-HIV-1 properties of SAMHD1. Retrovirology. 2013 Nov ​ 12;10(1):131. doi: 10.1186/1742-4690-10-131. PMID: 24219908 7. Halling DB, Kenrick SA, Riggs AF, Aldrich RW. Calcium-dependent stoichiometries of ​ the KCa2.2 (SK) intracellular domain/calmodulin complex in solution. J Gen Physiol, ​ 2014 Feb;143(2):231-52 8. Brookes, E, Perez, J, Cardinali, B, Profumo, A, Vachette, P and Rocco, M. Fibrinogen ​ species as resolved by HPLC-SAXS data processing within the UltraScan SOlution MOdeler (US SOMO) enhanced SAS module (2013) J. Appl. Cryst. 46:1823-1833 ​ ​ ​ 9. Sivagnanam S., Majumdar A, Yoshimoto K, Astakhov V, Bandrowski A, Martone M, and Carnevale T. Early Experiences in Developing and Managing the Neuroscience ​ ​ Gateway. Concurrency and Computation: Practice and Experience, issn 1532-0634, ​ 2014. 10. Brookes, Emre, Raminderjeet Singh, Marlon Pierce, Suresh Marru, Borries Demeler, and Mattia Rocco. "US-SOMO cluster methods: year one perspective." In Proceedings of the Conference on Extreme Science and Engineering Discovery Environment: Gateway to Discovery, p. 65. ACM, 2013. 11. Kanewala, Thejaka Amila, Suresh Marru, Jim Basney, and Marlon Pierce. "A Credential Store for Multi-Tenant Science Gateways." In Cluster, Cloud and Grid Computing (CCGrid), 2014 14th IEEE/ACM International Symposium on, pp. 445-454. IEEE, 2014. 12. Pierce, Marlon, Suresh Marru, Lahiru Gunathilake, Thejaka Amila Kanewala, Raminder Singh, Saminda Wijeratne, Chathuri Wimalasena et al. "Apache Airavata: Design and Directions of a Science Gateway Framework." In Science Gateways (IWSG), 2014 6th International Workshop on, pp. 48-54. IEEE, 2014. 13. Marru, Suresh, Rion Dooley, Nancy Wilkins-Diehr, Marlon Pierce, Mark Miller, Sudhakar Pamidighantam, and Julie Wernert. "Authoring a Science Gateway Cookbook." In Cluster Computing (CLUSTER), 2013 IEEE International Conference on, pp. 1-3. IEEE, 2013. 14. Pierce, Marlon; Marru, Suresh; Demeler, Borries; Majumdar, Amitava; Miller, Mark (2013): Science Gateway Operational Sustainability: Adopting a Platform-as-a-Service Approach. figshare. http://dx.doi.org/10.6084/m9.figshare.790760 Retrieved 18:03, Sep ​ ​ 16, 2014 (GMT) 15. Pierce, Marlon; Marru, Suresh; Mattmann, Chris (2013): Sustainable Cyberinfrastructure Software Through Open Governance. figshare. http://dx.doi.org/10.6084/m9.figshare.790761 Retrieved 18:04, Sep 16, 2014 (GMT) ​ Technologies or techniques (0 items) We made Apache Airavata releases​ 0.10-0.13 through the Apache Software Foundation during the reporting period.

Inventions, patent applications, and/or licenses (0 items) UltraScan Gateway: We have issued 309 user licenses for t​he UltraScan software (223 for ​ Microsoft Windows, 64 for Macintosh OS-X, and 22 for Linux), used in conjunction with the UltraScan gateway to pre-process, manage and visualize AUC data analysis jobs from the LIMS gateway.

Apache Airavata: Airavata software releases are licensed by the Apache Software Foundation ​ using the Apache License, Version 2.

Websites (0 items) ​ 1. http://scigap.org: general project information ​

Other products, such as data or databases, physical collections, audio or video products, software or NetWare, models, educational aids or curricula, instruments, or equipment (0 items) ​

1. https://github.com/SciGaP/: Project code repository. ​

Supporting Files

You may upload pdf files with images, tables, charts, or other graphics in support of this section. You may upload up to 4 pdf files with a maximum file size of 5 MB each. [TEXT GOES HERE]: Probably nothing needed here. Participants

For NSF purposes, for separately submitted and awarded collaborative proposals, the PI should report progress on his/her institution's portion of the collaborative effort only. In each of the subsections below, note which collaborators or contacts are involved in data contribution and/or management.

If there is nothing significant to report during this reporting period, please check "Nothing to Report", if applicable.

* Required fields What individuals have worked on the project?

Name Most Senior Project Role Nearest Person Month Worked

Marlon Pierce PI 3

Borries Demeler PI 6

Gary Gorbet Programmer 12

Mark Miller PI 2

Amit Majumdar Co-PI 2

Terri Schwartz Programmer 4

Paul Hoover Programmer 3

Kenneth Yoshimoto Programmer 2

Subhashini Sivagnanam, Programmer 2

Suresh Marru Co-PI 6

Saminda Wijeratne Programmer 10

David Reagan Programmer 2

Raminder Singh Programmer 3

Lahiru Gunathilake Programmer 6

What other organizations have been involved as partners? All Participants: XSEDE ​ UltraScan Gateway:Juelich Supercomputing Center (Unicore integration in Airavata, and ​ colocation services for LIMS server)

CIPRES Science Gateway University of Tennessee Knoxville (Integration of POPLar Gateway) ​ Have other collaborators or contacts been involved?

Some significant collaborators or contacts within the recipient's organization may not be covered by "What people have worked on the project?" Likewise, some significant collaborators or contacts outside the recipient's organization may not be covered under "What other organizations have been involved as partners?"

Impacts INSTRUCTIONS - This component will be used to describe ways in which the work, findings, and specific products of the project ​ have had an impact during this reporting period.

For NSF purposes, include, where appropriate, discussion of data resources and the acquisition of data skills. Include the emergence of new career paths, such as data scientists, or new disciplines.

If there is nothing significant to report during this reporting period, please check "Nothing to Report" if applicable.

Please make sure to read all instructions including NSF specific instructions, which can be found in the following link:

Required fields What is the impact on the development of the principal discipline(s) of the project? 8000 char limit. Describe how findings, results, techniques that were develo​ ped or extended, or other products from the project made an impact or are likely to make an impact on the base of knowledge, theory, and research and/or pedagogical methods in the principal disciplinary field(s) of the project.

Project Wide: By unifying the common features of three different gateways into a core set of ​ services, SciGaP will greatly improve the ability of all its client gateways to offer reliable, sustainable services.

Apache Airavata: during this project year, we made a significant investment of effort in ​ designing and implementing the API as well as restructuring many Airavata internal components based on the API. These enhancements are necessary for Airavata to work as a multi-tenanted, hosted services (as part of SciGaP’s overall goal).

UltraScan Gateway: In addition to the field of computer science, our contributions benefit the ​ field of hydrodynamics, especially investigators in biomedical applications, material science, nanotechnology and structural, functional, and molecular biology. The new methods available now through the UltraScan gateway provide unsurpassed detail in the analysis by leveraging high-performance computing through the gateway infrastructure. These improvements are particularly important to the data analysis of data from the new multiwavelength detector, because of the sheer data density that cannot be handled without parallel processing. Additional analysis methods investigate particle size and shape distributions, as well as density and hydrodynamic radius.

CIPRES Science Gateway: In addition to our main line contribution to computer science, the ​ project accelerates progress in virtually every field of biology, from Virology to Phylogeography.The methods made available through the CIPRES Gateway provide users around the world with access to parallel codes for computationally intensive sequence alignment and tree inference problems. This access both speeds the analysis of results, and also changes the landscape of what of which problems are computationally tractable for a given laboratory. The CIPRES Gateway was made resources available to more the 1100 institutions (Universities, Museums, Botanical Gardens, etc) in 73 countries on 6 continents, including 25 of 31 EPSCOR states/territories.

Neuroscience Gateway:In addition to contribution in computer science, the project has allowed ​ computational neuroscience researchers access to HPC resources via an administratively and technologically streamlined environment for uploading models, specifying HPC job parameters, querying running job status, receiving job completion notices, and storing and retrieving output data. This has allowed researchers to model high-dimensional parameter space exploration, study models that involve stochasticity, and simulate models with large neuronal networks and reproduce experimental results. NSG went into production in early 2013 and since then it has more than 200 users from about 20 different countries.

What is the impact on other disciplines? 8000 char limit Describe how the findings, results, or techniques that were developed or improved, or other products from the project made an impact or are likely to make an impact on other disciplines.

Science gateways in general support multidisciplinary work, enabling scientists in one domain to use tools from another domain that the find that they need for their research. The general purpose software and infrastructure we are developing in SciGaP, while a subject of cyberinfrastructure research itself, is also being built to provide an operational system that will be the platform for new science gateways in other scientific domains in later project years. What is the impact on the development of human resources? 8000 char limit Describe how the project made an impact or is likely to make an impact on human resource development in science, engineering, and technology.

Our approach to open source, open community development is already creating a pipeline of new developers for science gateways through the Google Summer of Code project as well as enabling and acknowledging the contributions of peer cyberinfrastructure developers. What is the impact on physical resources that form infrastructure? 8000 char limit Describe ways, if any, in which the project made an impact, or is likely to make an impact, on physical resources that form infrastructure, Including physical resources such as facilities, laboratories, or instruments. Science gateways are the infrastructure that provide science-centric views of scientific computing infrastructure. The impacts of the CIPRES Science Gateway and the emerging Neuroscience Gateway on bringing new users to XSEDE has been documented by the XSEDE project. The UltraScan gateway and its spinoffs have had a significant impact on the ability manage, analyze and understand data produced by scientific instruments. We expand on this below. Our challenge in the upcoming years of SciGaP is to extend these successes to enable new gateways providing access to new resources.

What is the impact on institutional resources that form infrastructure? Describe ways, if any, in which the project made an impact, or is likely to make an impact, on institutional resources that form infrastructure.

UltraScan Gateway:The developments for the UltraScan gateway significantly benefit core ​ facilities and individual laboratories because they allow parallelized data analysis of multiple experiments simultaneously. This dramatically reduces analysis time. The LIMS backend significantly improves accuracy and the high-resolution detail obtained from our parallel methods allow core facilities like CAUMA to provide better service and higher information content to their investigators.

CIPRES Science Gateway:Similarly, the CIPRES Gateway provides benefits to many ​ institutions, by allowing experiments to be conducted that could not easily be conducted with local institutional resources. The resource provided by access through CIPRES make individual analyses faster, and allow multiple analyses to be run simultaneously. The result is a significant savings in local infrastructure and system administration costs.

Neuroscience Gateway:The NSG provides benefits to many institutions where computational ​ neuroscience research is being pursued by graduate students and researchers dealing with complex and large neuronal models. These simulations require access to HPC resources and proper installation of optimized neuronal simulation tools on HPC resources and easy way to use the tools on HPC resources and retrieve output results. The NSG, by facilitating these, is allowing researchers from many institutions to carry out computational neuroscience research without having to make HPC resources and associated expertise available within their institutions.

What is the impact on information resources that form infrastructure? 8000 char limit Describe ways, if any, in which the project made an impact, or is likely to make an impact, on information resources that form infrastructure,

What is the impact on technology transfer? 8000 char limit Describe ways in which the project made an impact, or is likely to make an impact, on comme​ rcial technology or public use.

SciGaP’s Apache Airavata software is open source software that, like all Apache-licensed software, can be adapted to commercial use. Moreover, as an Apache Software Foundation project, Airavata is owned by the foundation rather by the SciGaP team members’ home institutions. This clarifies governance issues such as ownership of contributions and control of the project. In later project years, we will examine developing a business model around SciGaP as an Internet2 Net+ service. What is the impact on society beyond science and technology? 8000 char limit Describe how results from t​ he project made an impact, or are likely to make an impact, beyond the bounds of science, engineering, and the academic world.

By bringing open source, open governance methodologies to cyberinfrastructure development, we hope to provide better conduits between academic communities and commercial software development. By working in an open, public forums, we give students and junior programmers a chance to demonstrate their programming skills, problem solving abilities, and abilities to interact with a team. But we also are engaging in the other direction, bringing the thoughts and expertise of non-academic developers and architects to SciGaP problems through Apache Airavata’s architecture mailing list. Changes

INSTRUCTIONS -

The PI is reminded that the grantee is required to obtain prior written approval from the awarding agency grants official whenever there are significant changes in the project or its direction. See agency specific instructions for submission of these requests.

If not previously reported in writing to the agency through other mechanisms, provide the following additional information or state, "Nothing to Report", if applicable:

* Required fields Notifications and Request For more information on Grantee Notifications to and Requests for approval from the National Science Foundation, please visit the Notifications and Requests section in FastLane or refer to Exhibit II-1 of the Award and Administration Guide (AAG).

Changes in approach and reasons for change 8000 char limit ​

Actual or Anticipated problems or delays and actions or plans to resolve them 8000 char limit [TEXT GOES HERE]:

Changes that have significant impact on expenditures 8000 ​ char limit [TEXT GOES HERE]:

Significant changes in use or care of human subjects 8000 char limit [TEXT GOES HERE]:

Significant changes in use or care of vertebrate animals 8000 char limit [TEXT GOES HERE]:

Significant changes in use or care of biohazards 8000 char limit [TEXT GOES HERE]: