BioJava at at Google Summer of Code 2012

The Open Bioinformatics foundation as an umbrella organisation for
BioJava has been accepted to participate in this year's Google Summer
of Code. 

This means we will again be able to offer mentoring through BioJava
this year. Accepted students will get a stipend of 5,000$ from Google.
Participation is possible from most countries in the world, as long as
you are eligible to work in the country in which you'll reside
throughout the duration of the program.

If you are interested in working on a BioJava related project, now is
the time to start preparing and discussing your proposals. For the
last two years we had many applications for the projects proposed by
mentors. If you want to distinguish your application I recommend to
propose your own  project. Don't forget to discuss any proposal with
us before you submit them. We will try to provide feedback and match
you with a suitable Mentor.

Also see http://biojava.org/wiki/Google_Summer_of_Code and Google's
FAQs: http://www.google-melange.com/document/show/gsoc_program/google/gsoc2012/faqs

The student application deadline is April 6th. Google will announce
which proposals got accepted on April 23rd.

BioJava 3.0.3 released

BioJava 3.0.3 has been released and is available from
http://www.biojava.org/wiki/BioJava:Download as well as from the
BioJava maven repository at http://www.biojava.org/download/maven/ .


New Features

BioJava 3.0.3 adds several new features

- Significant improvements for the web service module (ncbi blast and
hmmer web services)

- Fastq parser (ported from the biojava 1 series to version 3)

- Support for SIFTS-PDB to UniProt mapping

- Improved support for working with external protein domain definitions

- Protmod module renamed to modfinder

- Numerous improvements all over the place (several hundred commits
since last release)

- We are also working on an update for the legacy biojava 1.8 series.

This release would not have been possible with contributions from
numerous people, thanks to all for their support!

Happy BioJava-ing!

RCSB PDB Fall 2011 web site update

Another major website update has been released for the RCSB web site. One of the main new features is Peter's new visualisation of surfaces with the Protein Workshop viewer. The calculation of surfaces is extremely fast and the resulting images look amazing. Here the capsid of a rather large virus (PBCV-1) - PDB ID 1M4X.

To visualize the surfaces, simply launch Protein Workshop from the viewer menu on the Structure Summary page.

Once the viewer has downloaded and installed itself, turn the surface on by dragging the Surfaces handle from Off to Opaque.

There are quite a number of display options, for a more detailed descriptions of all display options view here. 

A list of all the new features of the Fall 2011 release, is available on the "what's new page".

Google Scholar Citations Open To All

This week Google Scholar citations opened up to the public. After the initial release which was providing limited access only, the latest versions is now open for all scientists. What is interesting is that they seem to have created profiles for many scientists in an automated way. While the transparency of this is amazing, I can also imagine that there will be a push-back about privacy issues.

You can check out e.g who are the most cited scientists in Computational Biology or scientists at UCSD or check out Author profiles (e.g. yours truly). At the moment those lists are sorted by number of citations. Would be nice to have a few more sort fields like the already calculated h-index. Another missing feature is an API. Would be nice to be able to script their database and create custom reports.

I believe Google Scholar is setting a new standard for tracking success of scientists and is a serious threat to the ISI web of knowledge.

Google Summer of Code Mentor Summit 2011

This weekend I am spending at the Google Campus for the Google Summer of Code (GSoC) Mentor summit 2011. Just as in the previous years this is organized as an Unconference, a self-organizing conference without a prepared program. Instead the conference starts with the participants developing the program together. This year is the largest Mentor Summit ever and we are about 100 people over capacity (In total we are more than 300).

Hands up, "Who has edited a Wikipedia page?" almost every hand goes up. 

Some GSoC 2011 stats: 

This year Google accepted

48 new organisations 175 total orgs

1115 students

There were over 2k mentors

Students from 68 countries

Session Planning

If people want to organize sessions they have 30 seconds to say what topic they want to propose and write the title on a post-it note. The note will be put on a white board. Then people will vote on what sessions they want to attend. Finally the notes on the board are arranged according how the crowd likes.

After two minutes of planning a line is starting to form and many people are excited to propose sessions. Since so many people are proposing sessions the recommendation is to propose only one session per person.

Small selection of topics:

(too many to mention them all)

Wikipedia integration sucks.

How to build a million dollar computing infrastructure for free.

Open source (OS) gaming

How to pass the 5 developer barrier

OS for international development

LibreOffice: How to revive a project

Open Source for open science

How to build your own internet

GSoC What worked what did not work and how to improve next year.

Creating a cover archive for music

How to integrate semantic web into scientific applications

Visualisation of biological networks

Organizing the effort for documentation

How to get students to become long term contributors

Programmer oriented web semantics

Forming a non profit, how to fundraise for it

How to do telekinetic control of user interfaces

Refactoring the music industry

What to do if you have 100+ sub-modules

Community building

How to make the experience more local

How geeks give away their power with body language

How to get an keep female developers

Astronomy

Usability 

Open source in Asia

Aging project infrastructure

OS in higher education

How goverments can sponsor open source projects

Contribute to Openstreetmap

Detailed Program
http://gsoc-wiki.osuosl.org/index.php/2011
https://docs.google.com/spreadsheet/ccc?key=0AsVnlPsfrhUIdHZPVTlpMWJOZ3JwSHlFOURObEVMWnc#gid=0

Some Pictures


I like the built-in power plugs in the meeting room tables

The Google Tyrannosaurus

Session - Integrating  Creative Commons content

Demo: software to create music videos from Flickr.
Problem: nobody can get attributions right.

Flickr has 200 million CC images. However they are not all using the same variation of CC. Different types of CC are not compatible with each other.

now commenting on Etherpad:

http://etherpad.osuosl.org/integrating-creative-commons-content

Open Science Session
http://willyou.typewith.me/p/gsoc-open-science

Your Wikipedia Integration Sucks
http://etherpad.osuosl.org/IIAp0hzoid


Sunday


Cycling at the Google campus

Non-Profit Infrastructure for Software Freedom
http://etherpad.osuosl.org/im9y39OyFm

Fundraising 101 ("Free as in Freedom So Who Pays for the Beer?")
http://etherpad.osuosl.org/VK23qSKZCb
- plan
- don't depend on only one sponsor
- types of sponsors: indivduals, small businesses in community, businesses that use your software, local, foundations
-per-feature fundraising (-sometimes, be careful)
- What does a sponsor want to buy? - don't sell your email list. not everything is for sale. Virtue, in some cases forgiveness. Hiring
- And what are you prepared to sell? - ads on website, sponsored emails. Non-profits might have tax issues with certain types of ads. (depending on how they look)
- How to find sponsors. Ask everybody, look at your logs, use your mailing lists, be create, brave, hear no, be tenacious.
- Prepare your elevator pitch. Can you explain the benefit of your project in one sentence to an eight year old. Include your contact information. Make it easy to help. For events: plan and be realistic.
- Ask for help and offer help.
- Build a relationship, they might be able to connect you with somebody else who can sponsor.
- Make it easy to contribute, have a "donate now" button. Accept credit card, send thank you email.
- Enable Microdonations
- How to get graphic help. You need to go to where the artists are.


At lunch: a drone, remote controlled from a phone:

Walking around the campus

How to build your own Internet using open wireless

http://willyou.typewith.me/p/gsoc-openwireless

RCSB PDB web site September update

This week the latest update to the RCSB PDB web site went life. One of the major new features in this release is a new search interface. Alex has redesigned the top bar search box:

If you enter a search term, the new auto-suggest box provides suggestions what you might mean and allows to trigger precise searches, which are powered by an efficient new lookup mechanism in the background (written by Dimitris).

If you are not happy with the suggestions that you see (I can hardly imagine that ;-), you can still press enter and do the full text search across the PDB, which has been there already before.

Another major new feature of this release are the improvements to the PDB101, the educational section of the site. Proteins that are described in more detail as part of the Molecule of the Month articles, now have a Discussed Structure page. A particular nice detail that I want to point out is Greg's new 3D molecular viewer that works on iPhones and iPads. These gadgets usually can't display Jmol, our standard 3D viewer, due to the lack of Java support.  To work around this limitation there is now a new HTML5-sprite base animation. If you access a Discussed Structure page with a mobile device this is displayed and can be rotated left-right by "moving" it with your finger. Other improvements on the educational section are that the molecule of the month articles now show up in search results and you can download an article onto your phone. 

I will describe more new features some other time..

Tabular Report Web Services at the RCSB PDB

I have fallen behind on describing new features of the RCSB PDB web site... Let me catch up! One of the features that I find useful for people who want to script the site is that each of the queries on the site (and there are many) can be represented by simple XML.

How can you find that XML? Once you have run a query, you can access your query history from the left-hand menu. (We might actually move this to the top in the next release to make this more visible, but that's a different story.) Under the Query Details menu you can find the XML.



Here is an example XML for a query. This one below would simply list all current protein structures.

So what can you do with this XML?  You can POST this using the Search Web Services and get back a list of matching PDB IDs for this query. This functionality has been available for a while. As a new feature of the current RCSB PDB website release, Chunxiao extended this service to allow Custom Reports in a few different file formats.

Custom Reports


I have mentioned these reports already in a previous blog entry. On the Web site these reports allow to obtain Image Collages,  pre-defined reports containing various fields, and exports to Excel to mention just a few of the available options.

As part of the new Web service  these reports allow to fetch various fields via XML, comma separated file, or as an Excel file. There are step by step instructions available for how to use this new feature. Happy scripting!