-

Google Summer of Code

From Open Bioinformatics Foundation
Revision as of 13:53, 15 March 2009 by Chapmanb (talk) (BioSQL web interface and API on Google App Engine)
Jump to: navigation, search

The O|B|F is applying for the first time for the Google Summer of Code (GSoC) program as an umbrella organization for all O|B|F-affiliated projects.

On this page we are collecting ideas, possible projects, prerequisites, possible solution approaches, mentors, other people or channels to contact for more information or to bounce ideas off of, etc.

News

Contact

Our organization administrators are Hilmar Lapp (hlapp@gmx.net) and Mauricio Herrera Cuadra (mauricio@open-bio.org).

If you are a student interested in applying for a Google Summer of Code project with our organization, please send any questions you have, projects you would like to propose, etc to the developer mailing list of the pertinent O|B|F project.

How do you know which project is pertinent and the address of its developer mailing list? The projects under the O|B|F umbrella are listed below, with home page and developer mailing lists. Each project idea lists the O|B|F project it is a part of; look it up in the list below and you have the information you need. If you want to propose your own project idea and the project to which you would contribute isn't obvious, send email to gsoc@lists.open-bio.org.

Some of us also hang out regularly on IRC, see the list of O|B|F projects below for information on which projects have a channel and the name of the channel. (If you do not have an IRC client installed, you might find the comparison on Wikipedia, the Google directory, or the IRC Reviews helpful. For Macs, X-Chat Aqua works pretty well. If you have never used IRC, try the IRC Primer at IRC Help, which also has links to lots of other material.)

For applying, please make sure you read our documentation on information that students should know and guidelines we expect you to follow before you apply. We don't have a format template for application that you need to adhere to, but we do ask that you include specific kinds of information. What those are is documented under "When you apply."

Ideas

Note: if there is more than one mentor for a project, the primary mentor is in bold font. Biographical and other information on the mentors is linked to in the Mentors section.

Students: The below are only our project ideas, albeit well thought-out ones. You are welcome to propose your own project if none of those below catches your interest, or if your idea is more exciting to you, provided it is still a contribution to one the O|B|F member projects (see list below). Just be aware that we can't guarantee finding an appropriate mentor, but if we like your proposal we will try. Regardless of what you decide to do, make sure you read and follow the guidelines for students below.


Write a JEE5 webservice interface to BioSQL

Rationale 

BioSQL is a intelligently designed database schema for storing sequence data and associated metadata. It does however lack any kind of user API. A sensible way to design an API for a BioSQL backed database would be to expose the API as webservices. This would allow the API to be language and database agnostic (unlike an API based on database proceedures). It would also allow data in BioSQL to be very loosely coupled into bioinformatics workflows. Once an API is in place one could even adopt modified SQL schemas underneath as long as the data access API still conforms to some specification.

Approach 

Since the development of Java EE5 (and EJB3) the development of Enterprise Java Beans that interoperate with databases and webservices is exceptionally easy. In addition Java Session Beans can be readily exported as webservices with the addition of simple annotations, often no specific configuration is required. Free and open Java app servers (such as glassfish) that provide almost all of the management middleware for object relational mapping (ORM) and webservice deployment (and a whole host of other things) are available and relatively simple to use. Finally the free and open IDE Netbeans has excellent integration with Glassfish and Java EE5 (plus I am most experienced with this IDE so I can provide more help with it's use). For these reasons I would suggest that Java EE5 is the most sensible approach to implementing this project.

During a development meeting, in Tokyo in 2008, a preliminary EJB mapping to BioSQL was generated. What remains to be done is the development of a simple, well documented and well tested API specification and implementation that bioinformatics developers can use to perform CRUD (CReate, Update, Delete) functions on the database as well as useful search and retreival operations.

In summary the project will define and document an API and expected behaivour and then implement the webservice interface. A set of unit tests will also be developed along with a proof-of-concept app that demonstrates use of the API.

Challenges 
  • Designing and documenting the API so that it is simple and intuitive
  • Making simple queries simple and efficient and complex queries possible.
  • Making CRUD operations secure (only people with the right credentials should be able to delete the data).
  • Loaders for common file types.
  • [Nice to have] Making a test application that will call API methods with predefined arguments. This will let people make alternative implementations of the API while testing they are still compatible with the API. For example someone could make an entire implementation in Perl/ BioPerl and still have it validate against the API.
Involved toolkits or projects 

JavaEE5, BioSQL, parts of BioJava would be useful to steal for parsing.

Degree of difficulty and needed skills 

Medium to Hard. While the use of Java EE5 is now quite easy (esp with IDEs like Netbeans) there is quite a lot of concepts involved in the project (Webservices, ORM, EJBs etc). The hard part would be getting up to speed with those concepts. If you already know a lot of this then the project would only be medium difficulty. At minimum the student should be confident with Java and at least aware of some of the technologies. This is not the right project for a very new programmer.

Mentors 
Mark Schreiber (and anyone else who wants to help)

Mapping the NCBI toolkit to BioPerl, BioRuby, BioConductor and BioJAVA using BioLib

Rationale 

The National Center for Biotechnology Information (NCBI) has created a large collection of utilities developed for the production and distribution of GenBank, Entrez, BLAST, and related services. To support these utilities a large set of C and C++ libraries are maintained and regularly improved by NCBI. These include, for example, sequence alignment algorithms, antigenic determinant prediction, CPG-island finder, ORF finder and string matchers. This functionality is ultimately of great interest to all scientists working in molecular biology with application in biology and biomedical research.

Unfortunately, few bioinformaticians work with C/C++. Addressing this NCBI has made a binding available for Python. This is not enough as bioinformaticians work in many different programming languages, and to be fully effective support should be made available at least for Perl, R and JAVA. These three together, probably, representing over 90% of bioinformaticians. The BioLib project successfully provides the 'mapping' infrastructure to map complex libraries against many computer languages using SWIG. Basically one mapping suffices to support all popular languages.

Approach 

Special interfaces need to be developed to map the NCBI toolkit libraries against Perl initially. The (outdated) NCBI Python mapping can be used as an initial guide for mapping functionality. Once mapped against Perl mapping against Ruby and Python is trivial. However, at this point BioLib support for R and JAVA needs to be developed. A proof-of-concept can be part of this project. Finally SWIG mappings can be used to create automated documentation and testing of BioLib code.

Challenges 

The main challenge is to provide nice and consistent interfaces in high-level languages against the NCBI C/C++ toolkit library. This requires OOP design and unit testing of existing functionality. Also some SWIG hacking may be involved to provide decent mappings for R and JAVA, as well as SWIG auto generated documentation and testing.

Involved toolkits or projects 

BioLib, BioPerl, SWIG (and optionally BioRuby, R/Bioconductor, BioJAVA or BioPython)

Degree of difficulty and needed skills 

This is a challenging project as it crosses computer languages. It requires experience in C++ and a wish for deeper understanding of at least one high-level OOP language like Perl (did I write OOP?), Python, JAVA, R or Ruby.

Mentors 

Pjotr Prins, Chris Fields

BioSQL web interface and API on Google App Engine

Rationale 

The BioSQL project provides a robust and well supported database schema for storing sequence data and associated annotations and features. It does not have a standard web interface or web facing API, both of which would provide improved access to scientific data. Deployment of BioSQL currently requires knowledge and administration of relational databases, which can hinder its use in smaller research laboratories that do not have public servers or experienced systems administrators.

This proposal seeks to bridge this gap by providing a rapidly deployable cloud based solution utilizing the established BioSQL backend. This system will allow scientists to share results in a standard format both early on during research and at the time of publication. By deploying on stable architectures, long term data access is ensured and not dependent on maintenance of local servers. Data archival for replication and expansion of ideas is an important part of the scientific process; this recent blog review summarizes some of the problems associated with primary data access.

Approach 

Google App Engine provides a full development stack for rapidly building and deploying web applications. The platform provides free quotas which allow a small lab with a limited budget to make their data available, and also scales for larger projects with popular data sets.

The student project expands an initial demonstration server (demo server; source code; blog post) to a full featured web application. The server side implementation will be programmed in Python, utilizing the Google App Engine developers toolkit supplemented with the Biopython libraries. The client web interface will be designed using HTML, CSS and javascript; the interface will utilize a full featured javascript library, such as jQuery and jQueryUI or ExtJS. Client to server communication occurs using AJAX techniques with JSON for data exchange.

In addition to the web interface, the server will also provide a programming interface using a REST API. This involves coordination with other proposed projects, including the proposed JEE5 Java webservice, to design a common interface.

Challenges 
  • Familiarizing student with Python, Javascript and AJAX, as well as the Google App Engine environment.
  • Initial implementation of BioSQL server interface with useful features.
  • Coordinating input from users on the BioSQL mailing list. The student will need to solicit desired features from users and prioritize based on implementation time and importance. See this mailing list discussion for an example of interest and initial ideas.
  • Designing the web interface for intuitive use.
  • Coordinating API development with other projects.
Involved toolkits or projects 
Degree of difficulty and needed skills 

Medium to Hard. This requires a familiarity with current web frameworks and utilizes a number of existing libraries to allow the student to jump right into the development process. This requires the interested student to be comfortable with quickly learning outside libraries. Beyond programming, the project will also involve creative thinking about interface and usability design.

Mentors 

Brad Chapman (plus...)

Biogeographical and community phylogenetics for BioPython

(Note: this project is proposed by potential GSoC student Nick Matzke.)

Rationale 
The field of phylogenetics has proliferated, and one new development is that large, phylogenetically explicit datasets are beginning to be used to answer questions about the relationships of ecological communities and biogeographic regions, instead of just individual clades. The phylocom package (Webb et al., 2008) contains fast C implementations of basic analyses such as alpha- and beta-phylodiversity (Net Related Index and Nearest Taxon Index). The R package picante, funded by NESCent and Google Summer of Code 2008, contains utilities for processing phylocom inputs/outputs as well as additional tools for applied phylogenetics such as phylogenetic signal, phylosor (phylogenetic sorenon's index), and lineages-through-time plots. These tools, developed for evolutionary community ecology, are useable in any context where a collection of lineages are undergoing cladogenesis, dispersal, and extinction in a series of containers (communities, biogeographic regions, gene families undergoing gene conversion, laterally transferring elements in unicell genomes, etc.)
The related field of phylogenetic or historical biogeography -- the estimation of the geographic location of ancestral lineages, the history of their dispersal, and the history of connectivity and vicariance between regions -- has also advanced with a variety of algorithms (Ronquist's Dispersal-Vicariance Analysis, DIVA; lagrange, a maximum likelihood method implemented in Python, available online at Google Code; GeoPhyloBuilder, a NESCent-sponsored package for producing GIS files to display biogeographic history in Google Earth; croizat, a panbiogeographical method and visualization package implemented in python using matplotlib's Basemap module; and older methods derived from traditional ancestral-state reconstruction).
Approach 
Write BioPython modules/functions to:
("*" indicates some version of this already done independently by Nick Matzke)
  • Improve BioPython's Bio.Nexus.Trees newick parser, which currently cannot successfully read the newick files output by Phylocom (although these files are read successfully by a variety of other programs and modules, e.g. Dendroscope, alfacinha python module).*
  • Implement Cardona et al.'s Extended Newick format for reticulating trees etc. (only exists in BioPerl currently)
  • Develop a series of functions for processing phylocom inputs and outputs*
  • Provide functions for basic community/geographic relatedness (e.g., NRI, NTI, phylosor)*
  • Calculating these statistics for large phylogenies requires calculating/processing a large distance matrix with a C or java library*
  • Basic graphics for analyzing community/regional phylogenetic history, e.g. lineage-through-time plots*
  • Downloading sample location data from online databases (e.g. GBIF, although see here), combine with phylogenies for input into lagrange, DIVA or other algorithms
  • Re-creating DIVA in Python; the only available version is 12 years old and currently will only run on certain PCs
  • Process output from DIVA, lagrange, etc., for display in GISs, Google Earth (KML files), and/or matplotlib's Basemap
Challenges 
  • Contacting & involving/getting feedback from authors of the mentioned packages (have been in contact with many of them already)
  • Uncertainty, error, & missing data in geographic location databases (see here), and flagging such
  • Deciding the appropriate number of BioPython modules, etc. will require mentor advice
Involved toolkits or projects 
Degree of difficulty and needed skills 
Medium. Requires a familiarity with not just python/biopython but some unusual data formats and datasets, and packages, and integrating them (geographic, phylogenetic, metadata, etc.). Must be familiar with evolution, phylogenetics, biogeography, and the statistical hazards from oversimple interpretations of these.
Mentors 
Brad Chapman (plus? Various python/phylogenetics gurus at NESCent etc might be consulted)

phyloXML support in BioRuby

Rationale 
Evolutionary trees are central to comparative genomics studies. Trees used in this context are usually annotated with a variety of data elements, such as taxonomic information, genome-related data (gene names, functional annotations) and gene duplication events, as well as information related to the evolutionary tree itself (branch lengths, support values). phyloXML is an XML data exchange standard that can represent this data. Trees in phyloXML format can be displayed and analyzed with Archaeopteryx (the successor to ATV), which also allows manipulation and navigation of the tree. While tools exist to convert other formats (such as the widely used Newick and Nexus formats) to phyloXML, there is currently support for phyloXML in only one of the open source Bio* projects (in BioPerl, as a result of Google's Summer of Code 2008).
Approach 
Build phyloXML support in Ruby. More specifically, extend the open source BioRuby project to support phyloXML (BioRuby 1.3.0 has just been released). This will entail (i) the development of objects to represent all the elements of phyloXML (sequences, taxonomic data, annotations, etc), (ii) the development of a parser to read in phyloXML, and (iii) a phyloXML writer.
Challenges 
Relating the data elements specific to phyloXML to the tree classes already in BioRuby while maintaining the standards of the BioRuby project. Development of a time and memory efficient phyloXML parser (the parser has to be able to process trees with thousands of external nodes, at least).
Involved toolkits or projects 
BioRuby, phyloXML
Degree of difficulty and needed skills 
Medium. Requires experience in an object oriented programming language (such as C++, Java, or, ideally, Ruby). Experience in genomics or a related biological field is also critical. Knowledge of BioRuby will obviously help, as well as familiarity with XML.
Mentors 
Christian Zmasek (and anyone else who wants to help)

BioPerl integration of the NeXML exchange standard + Bio::Phylo toolkit

Rationale 
NeXML is an emerging XML standard for the serialization and exchange of phylogenetic information. In Perl, the Bio::Phylo toolkit is the preferred parser/writer interface for NeXML. While Bio::Phylo contains methods that will operate on BioPerl objects [such as alignments (Bio::SimpleAlign) or trees (Bio::Tree)], a set of methods to wrap Bio::Phylo functionality into BioPerl in a systematic and updateable way would lower barriers to broader use of this useful standard.
Approach 
We would like to explore a couple of ways to form the linkage between BioPerl and Bio::Phylo, while still maintaining Bio::Phylo's independence as a module. Since it is part of the implementation side of a rapidly evolving standard, it is more mutable than the average BioPerl module, and should be more nimble. One method would be implement a thin BioPerl wrapper around Bio::Phylo, that allows BioPerl objects to be passed easily in and out, and maintains a stable BioPerl-compliant API, hiding Bio::Phlyo API changes. However, since this project is exploratory, we could also prototype a version of Bio::Phylo that is directly implemented as a BioPerl module. We would also develop appropriate usage tests, test data sets, target audience use cases, benchmarks and profiles to compare the approaches we come up with.
Challenges 
  • Designing a relatively stable wrapper around a relatively dynamic module;
  • Designing tests that cover important use case scenarios meaningful to BioPerl users;
  • Identifying and interfacing Bio::Phylo output and NeXML-serialized data with up- and downstream BioPerl operations; e.g., adding a Bio::SeqIO::nexml module for doing BioPerl-native NeXML IO.
Involved toolkits or projects 
Degree of difficulty and needed skills 
Easy to medium difficulty. Perl fluency required; experience with object-oriented Perl very helpful; experience with biological data (sequences, sequence alignments, phylogenetic trees) a plus; experience with BioPerl itself will flatten the learning curve.
Mentors 
Mark Jensen, ...(rvos?),...

Mentors

What should prospective students know?

Before you apply

  • If you want to apply with your own idea, determine which O|B|F project you would be contributing to, and contact us early on so we can try to find a mentor.
  • Our scope for proposals that we will entertain is those extend one of affiliated toolkits. Project proposals that would create a new stand-alone piece of code are outside of our scope.
  • We are most interested in students who give us evidence that they have already or might develop a sustained interest in becoming future contributors to one (or more) of our projects.
  • Ask us questions about the project idea you have in mind.
  • Write a project proposal draft, include a project plan (see below), and bounce those off of us.


Have I mentioned yet that you should be in touch with us before you apply? The value of frequent and early communication in contributing to a distributed and collaboratively developed project can hardly be overemphasized. The same is true for becoming part of a community, even if only temporarily.

When you apply

When applying, (aside from the information requested by Google) please provide the following in your application material.

  1. Why you are interested in the project you are proposing, uniquely suited to undertake it, and what do you anticipate to gain from it.
  2. Why are you interested in contributing to the O|B|F project that your work would be (or become) a part of? To what extent and in which ways do you anticipate to stay involved with the project?
  3. A summary of your programming experience and skills.
  4. Programs or projects you have previously authored or contributed to, in particular those available as open-source, including, if applicable, any past Summer of Code involvement.
  5. A project plan for the project you are proposing, even if your proposed project is directly based on one of the ideas above.
    • A project plan in principle divides up the whole project into a series of manageable milestones and timelines that, when all accomplished, logically lead to the end goal(s) of the project. Put in another way, a project plan explains what you expect you will need to be doing, and what you expect you need to have accomplished, at which time, so that at the end you reach the goals of the project.
    • Do not take this part lightly. A compelling plan takes a significant amount of work. Empirically, applications with no or a hastily composed project plan have not been competitive, and a more thorough project plan can easily make an applicant outcompete another with more advanced skills.
    • A good plan will require you to thoroughly think about the project itself and how one might want to go about the work.
    • We don't expect you to have all the experience, background, and knowledge to come up with the final, real work plan on your own at the time you apply. We do expect your plan to demonstrate, however, that you have made the effort and thoroughly dissected the goals into tasks and successive accomplishments that make sense.
    • We strongly recommend that you bounce your proposed project and your project plan draft off of us, using either the pertinent developers mailing list or the IRC channel(s). Through the project plan exercise you will inevitably discover that you are missing a lot of the pieces - we are there to help you fill those in as best as we can.
  6. Your possibly conflicting obligations or plans for the summer during the coding period.
    • Although there are no hard and fast rules about how much you can do in parallel to your Summer of Code project, we do expect the project to be your primary focus of attention over the summer. If you look at your Summer of Code project as a part-time occupation, please don't apply to us.
    • That notwithstanding, if you have the time-management skills to manage other work obligations concurrent with your Summer of Code project, feel encouraged to make your case and support it with evidence.
    • Most important of all, be upfront. If it turns out later that you weren't clear about other obligations, at best (i.e., if your accomplishment record at that point is spotless) it destroys our trust. Also, if you are accepted, don't take on additional obligations before discussing those with your mentor.
    • One of the most common reasons for students to struggle or fail is being overstretched. Don't set yourself up for that - at best it would greatly diminish the amount of fun you'll have with your Summer of Code project.

Other information

Open-Bio projects involved

BioPerl logo tiny.jpg
BioPerl 
Biojava logo tiny.jpg
BioJava 
Biopython logo tiny.png
Biopython 
BioRuby logo tiny.png
BioRuby 
BioSQL logo.png
BioSQL 
BioLib logo tiny.png
BioLib 

Reference Facts & Links: Google Summer of Code 2009

GSoC2009Logo.png
  • Mentoring organizations apply between March 9-13, 2009. Accepted mentoring organizations will be published March 18. See full set of timelines.
  • Google expects to accept around 150 mentoring organizations, a bit less than in 2008 (when they accepted 175). If the trend over the past years is any indication, this will be out of at least 3x as many organizations that apply.
  • Students apply between March 23-April 3, 2009. The eligibility requirements for students are in the GSoC FAQ.
  • Development occurs on-line, there is no requirement or expectation to travel, neither for students nor for mentors.