As you know, my post-doc in Uppsala ended. It was\na good time, and it was great collaborating on Bioclipse with Ola, Jonathan,\nArvid, and Carl. I would have loved tighter integration with the work of Maris\nand Martin, but that was limited to one joined paper (in press). I thank Professors\nJarl Wikberg and Eva Brittebo for allowing me to continue my research at their\ndepartment, and hope this is not the end of the collaboration yet.
\n\nLike with new year, the end of a contract is a good time to reflect on ones\naccomplishments. It’s been a bit delayed, but as you know, I already in my next\nproject in Cambridge ,\nand will start in January with yet a longer term position in predictive toxicology\n(more on that soon). This makes this a really crowded period, on top of birthdays,\nSinterklaas, x-mas, and sorts.
\n\nMy Research
\nAs you might know, my research interest lies in understanding molecular properties\nand their applications in larger molecular systems. This can be how small molecules\npack in crystals, finding patterns in properties (QSAR-like work), etc. Because\nthe underlying methods are useful in many domains, you see applications in various\ntoo, including drug discovery, metabolomics, etc. These methods involve statistics\nand cheminformatics, primarily, which is clear from my publications on method\ndevelopment in chemometrics and cheminformatics. You will also have seen that\nvisualization is a very important tool here, as our numerical validation can\neasily mislead even a trained scientist.
How Uppsala fits in
\nAbout 30 months ago, I got an offer to join the Bioclipse\nteam to work on the cheminformatics features of the workbench. It was already\nusing the CDK, so the project was a tight match with what\nI did in the past. Additionally, there were plans to integrate R,\nand while the latter is partially implemented, that part was unfortunately not\ncompleted by the group yet. I believe this is a crucial aspect, and without\nit the large-scale impact of Bioclipse will be severely reduced.
Bioclipse is positioned as a workbench to use third-party libraries, web services,\ndatabases, etc, and has done so very successfully (doi:10.1186/1471-2105-8-59).\nIt speaks many Open Standards, and already incorporates various important Open\nSource libraries for life sciences research, including the aforementioned CDK\n(doi:10.2174/138161206777585274,\ndoi:10.1021/ci025584y), but also Jmol,\nJChemPaint, BioJava (doi:10.1093/bioinformatics/btn397),\nand others. Using these libraries it has rich visualization means for life sciences\ndata, including molecules and protein sequences. The latter, of course, is directly\nrelated to the proteochemometrics research done in Wikberg’s group.\nRecently, Bioclipse adopted scripting functionality, making it a perfect tool\nto share life sciences computation, just like Taverna (doi:10.1093/bioinformatics/bth361)\nand KNIME.
\n\nWhere I hoped to do some research in proteochemometrics, events lead me into\ndifferent areas, which I explained in Why you have not heard me much about\nchemometrics recently… .
\n\nBut, Bioclipse provides me with the tools I need to take molecular chemometrics\n(doi:10.1080/10408340600969601)\nforward.
\n\nResults
\nSo, what has this resulted in, besides a number of unsuccessful \ngrant applications? We’re still counting, but two book chapters, a book on pharmaceutical\nbioinformatics, one proceedings paper, five research\npapers, seven oral presentations at international meetings, and a ACS conference\non RDF in chemistry. Oh, and tons of\nOpen Source code, of course. (I’m at the edge of collapsing; I did that as student,\nlost a year, but learned a lot about myself … these results I have worked\nvery hard for; I am not a miracle worker.\nAnd I have to disappoint people occasionally, as things do not work out how\nI expected them to be. My apologies for that.)
I will not describe all in detail now, but focus on a few things around what\nI made my research in Uppsala: semantic cheminformatics, which I believe to\nbe a key concept of where cheminformatics must be going. The first paper resulted\nfrom a collaboration with Johannes, a medical researcher at the Ludwig-Maximilians-Universität\nin Germany (full reference at the end of the post). This work provides an alternative\nto SOAP, which has a better solution to asynchronous computing that the polling\napproaches now commonly used. A XMPP-based service just reports back when it\nis done, so that you do not have to ask all the time. Makes sense to me. We\nmade the platform available to Bioclipse and Taverna, and demonstrated the technology\nwith applications in life sciences, including (QSAR) descriptor calculation,\nand susceptibility for seven known HIV protease inhibitors.
\n\nThis work stresses that if we really want to, we can significantly improve scientific\ncomputing. It’s very much like what Peter concluded this week:\n“None of this is rocket science - it’s purely a question of will”. This is what\nI have being trying to show in the past few years. The disuse of accurate scientific\ncomputing is a deliberate choice. Making your cheminformatics research irreproducible\nis a choice, and a bad one too. There can be acceptable reasons, but the choice\nwould be bad nevertheless (I hope that distinction is clear: you can have valid\nreasons to do something intrinsically wrong. You will be forgiven, and be encouraged\nto change your behavior.) Many people on the Blue Obelisk community are laying\nout the foundations and show cases, hoping to make it easier for others to change\nbehavior. I think we have been quite successful there.
\n\nAnyways… on to the second paper. As said, Bioclipse is the platform that can\nbring these new cheminformatics methods to the desktop. The new and improved\nBioclipse 2 (see citation below) adds one important new feature: scripting.\nMy work in this paper focuses on doing making sure the cheminformatics library\nwas properly integrated, continued development of JChemPaint (yet unpublished,\nand in collaboration with the EBI, but see\nfor example this blog post ),\nand helping Ola and other to properly use the CDK in their applications (MetaPrint2D\n(doi:10.1186/1471-2105-11-362),\netc.). The impact of this work goes far beyond the papers on which I am author,\nthough not every reviewer will understand that, unfortunately. This work is\nreally the plumbing, it’s the development of the measuring machines to do the\njob, the development of a STM device to actually get going.
\n\nThe third paper, also listed at the end, is about defining a standard for detailed\nexchange of QSAR data. It defines what information is needed to reproduce a\nset of QSAR descriptors, including the input, and using a descriptor ontology\nwhich we published about before (doi:10.1021/ci050400b).\nThis project can be the seed of a public repository of QSAR data, where it will\nbe clear what is meant, and how the data can be used. If you are interested\nin setting up such a public repository, please contact me or Ola.
\n\nThat leaves me to the work that I have initiated in the group: the use of RDF\ntechnologies (I\ndo hope all VR reviewers are listening). RDF provide a lingua franca\nfor data exchange in life sciences, and the meaning of words is provided by\nsharing dictionaries (ontologies). Bioclipse has been extended to speak RDF,\nand we developed various applications based on it. A proceedings\npreviews the effort, while the paper is in print in the new Open Access Journal\nof Biomedical Semantics. Of course, you can also\nread much about this topic in this blog.
\n\nRDF is going to change bio- and cheminformatics in ways the XML has been unable\nto do. Various papers are currently in preparation to provide detailed uses\ncase and related research. I am very excited about this technology which further\nimproved interoperability and reproducibility in cheminformatics. Should you\ncare about that? Yes, because by using these good practices, research will be\neasier to interpret, conclusions judges, and as such, we can focus on the underlying\nchemistry in much more details, instead of looking at noise which many current\ncheminformatics literature is doing. (Ouch, that’s a bold statement indeed.\nTrue? Well, without reproducibility it is hard to tell. Let’s all work towards\nless magic, less black box, and more science in this field; we will all benefit\nfrom that. Who knows, we might even convince the bench chemist that we are doing\nsomething right ;)
\n\nSo, where is the understanding of underlying patterns, you may wonder? That\nis a fair question, but I have no grudge in admitting that after my PhD that\npart has been underrepresented. That will change soon enough, though. Now I\ncan only hope it is on time go get me a Nature or Science paper, required to\nget tenure (see this discussion).
\n\nThat’s not all I did. I have not discussed the book chapters, the book, the\nother publications to which I contributed in various ways (doi:10.1186/1471-2105-11-159,\ndoi:10.1093/bioinformatics/btq476).\nThat will come in a more detailed report later.
\n\nFinally, I link to thanx Uppsala University for the KoF\n07 grant which funded my work in Uppsala.
\n\n