Reflection on Open Science Workshop @ PSB 2009 (1)

Open Science Workshop is what brought me to PSB 2009 (Pacific Symposium on Biocomputing) and what urged me to start this blog for recording my scientific life. Thanks again for Cameron and Shirley for organizing such a great event and helping with my travel expense from funds provided by the Burroughs Wellcome Fund (in addition to the travel award from Research and Conference Grants from University of Hong Kong). Thanks also in particular to Shirley for all the efforts in making this open science workshop open. All materials related to the workshop can be found in friendfeed room, slideshare and an unoffical offical blog.

I shared a similar path as Cameron did to get into open science, and thought of similar ideas such as creating a LaBlog, and having workflow being built on any experiment. While Cameron has actively experimented a number of possibilities, my idea still remains as an abstract and is moving towards creating my own platform for data recording, exchange, analysis, publishing and sharing. I tend to work for the flipped side of openness (as described by Shirley). I am not keen on encouraging openness in the social construct but envision technology would make science easy for researcher where any data could be easily browsed, linked and shared with any data-intensive hypothesis could be easily queried and corelated. The concept of one big lab should come effortlessly and optionally. I will elaborate more later.

The workshop has been summarized nicely here.  So, I just focus on my reflection on the workshop.

Cameron first talked about why he cared if science is open. He mentioned doing research is a privilege and we should ensure every dollars spent are worthwhile. More importantly, making science more open can avoid having many post-doc repeated the same experiments which resulted in both a waste of time and money. I agree, for the first point, that’s undeniable (or cannot be denied publicly ). However, for the latter, who cares? We all know that even for published data, the experiment may be hardly reproducible. So, for every first step of approaching a question, we need to repeat others experiment first to verify if that’s real and correctly described. So, rather than avoiding repeats of experiment, I think open science allow results be better documented such that a bigger or more complete picture can be drawn from many independent experiments. The main hurdles of open science, in my opinion, would be the willingness and ease of sharing. Traditionally, scientists work for a problem which they interested in, and kind of enjoying the ’self-employed’ status. They enjoy the path of searching and attaching their name to ‘answers’ or ’solutions’ but seldom link it to its application nor contribution to the society. Though it’s going to be changed for the upcoming more difficult problems that require collaborative efforts around the world across multi-disciplines (e.g. collaborative cross in mice and HapMap in human), the latter become the major limiting step of achieving open science. I always think people are keen on sharing, but not only until facebook was launched, we do not know it’s the technology that limits the sharing ability of people. and therefore, I am inclined towards improving the technical side of open science. Is it currently easy enough for researcher to put up their data on the web and do simple query? Obviously not, but the community is certainly working very hard here.

Philip Bourne, the keynote speaker of the workshop, is one of them. After having found the Plos Computational Biology Journal, Phil continues to improve the functionality of a journal by creating SciVee and BioLit. I do agree with his viewpoint of which in the future a biological journal should be indistinguishable from a biological database. Users should not be required to copy a sequence from a paper and paste it onto NCBI blast engine or download the genotype data of the Japanese from HapMap and generate the haplotype structure with HaploView in order to compare it to that of your data collected from Southern Chinese. In my recent years of being a student in a molecular biology lab, I agree very much on his words, ‘lots of common exercises [nowadays] are beyond of the work of a biologist.’ therefore, rather than encouraging the biologist to adapt the web environment, and put extra efforts in scanning, copying, pasting and formating their lab workflow onto the web, we require ourselves, as a bioinformatician, to develop tools that allow easy export or even ultimately run natively on the web inter-connecting different services and blogs.

and in fact, I do know informatics people are moving towards this direction. David DeRoure and Carole Goble mentioned their project in UK, myExperiment, aiming to let researchers share their scientific discovery workflow. I believe it provides a necessary backbone or platform for future development of bioinformatics tools. They also recognized the coming explosion of bioinformatics web services from the community, and thus initiated a project called BioCatalogue, to “provide a single registration point for Web Service providers and a single search site for scientists and developers”.

P.S. Speed on the web is really fast. While I am writing this post, lots of posts about the workshop had been put up. Somehow, it’s a pressure or even discouragement to a new blogger, but anyway, I think it’s better for me to catch up that speed, coz that’s all the web offers — effective and efficient.

P.S. To me, it’s rather a long post, however, only half of the time I used for writing itself. Another half was spent on finding relevant links, doing all kinds of copy and paste, and being discouraged by the unable network. Writing a blog does not as easy as it seems.