Monthly Archives: June 2006

The Hapmap and personalized medicine

The October 2005 issue of Nature had the complete results of the first stage of the Hapmap project. I first came across the Hapmap when I heard Chris Smiths Nature podcast corresponding to that issue. The first stage marked the analysis of polymorphisms in 270 individuals who came from diverse ethnic background. Although I am no geneticist. The great thing about the hapmap data is that it finally allows us to understand human genetic variation.

We have all heard that the genetic differences between all of us is very very small. Of course the hapmap is basically a way to understand what are those differences and how small is very small. Perhaps one of the emerging facts , from studies leading up-to the hapmap project , was the extent of linkage dis-equilibrium among our differences. I will not even attempt to explain linkage-disequilibrium: suffice to say that that disequilibrium arises from differences traveling together from one generation to the next. i.e they are linked. Now why is that a big deal ?. Well it means that instead of studying the 7 million odd “polymorphisms” ( i.e the differences between you and me). Thanks to linkage disequilibrium you can possibly get at a good picture of how “generally different ” you are from me by looking at around just 10,000 places.

SO why is this big deal, because it drastically reduces the cost and effort of tracking down these differences.

Now think of a drug undergoing clinical trials. We have all heard of people who remark ” I dont respond well to amoxicillin but do just fine with zithromax ( two different antibiotics). Now imagine the scenario during a clinical trial , when that “you” and “me” is now linked in with your haplotype ( the collection of differences) and my haplotype and then at the end they might just have a clear picture of – People with haplotype X cannot tolerate this drug while people with haplotype Y , can. That is powerful information that can potentially save lives.

Also the hapmap and the study of such polymorphisms also has powerful implications for the study of human anthropology , human history and human natural selection. In this context the online seminar from science magazine explaining the process of using SNP data to examine natural selection makes very interesting viewing ( see References).

Overall, there is no denying the impact of rapid genotyping on understanding and classifying human response to clinical trials. Of course , just as the launch of Bidil -the first pre hapmap era “personalized drug” specifically targetted towards African-Americans set of a whole slew of ethical debate. I am positive that having hapmap data guide drug discovery or dosage may also be much of a mixed blessing in this era of genomic-medicine

Relevant Links :

Clues to Our Past: Mining the Human Genome for Signs of Recent Selection

Online Science Magazine flash seminar on evidence for positive selection in human lineages and the review by Sabtei et. al.

Race based drug may halt trial for better clinical markers

Nature news and Views on the Hapmap project

Wikipedia entry with an excellent description of what is a haplotype and an SNP

Tag spam?? what will they think of next!

I just started using connotea a del.icio.us clone from the Nature magazine people that brings social bookmarking and tagging to scientific citations.

Connotea makes it very easy to tag and store all online material and it has a powerfull doi lookup feature whereby just highlighting a doi (digital object identifier) and clicking on the Bookmark “Add to connotea” on the browser toolbar , adds a full citation record to your connotea library. While adding such a citation you can “tag” it as appropriate for your and others future use and retrieval.

Connotea also supports OpenURL appending which validates any licenced content with your library server if you set it up and then serves it up seamlessly. On the whole, with the doi based entry and ability to store tagged collections it is a very handy tool for anyone who reads a lot of papers online.

And now for the annoyances. I wanted to look at a sampling of references or bookmarks with the tag “Bioinformatics” on connotea to discover new material from other users. So, using the search feature, I looked for entries with the global tag “Bioinformatics “. I was then shocked to find that the top 20 or so matches coressponded, not to serious or even remotely usefull science or other website links, but to what was obviously advertisements from a drug vendor website whos name I will not mention here because it will give them some more un-necessary publicity.

This seemed like a desperate attempt by the said website to hijack the connotea system to put in links for their own products which had nothing to do with bioinformatics.

Fortunately, I have not seen this on del.icio.us yet. But this Tag-spamming is indeed an unfortunate development which could seriously pollute the web2.0 social-network.

Addendum: Kudos connotea

After writing to connotea support about this problem. I got a very prompt email informing me that all links from the offending user have been removed and that they are working on improving various automated and user driven procedures to weed out spammers. Their quick and detailed response has me greatly impressed.

I have also started using the “Add to connotea” greasemonkey script from Piere Lindenbaum . And saving pubmed searches to connotea is all very easy now.

Nature network Boston

Only recently did I start reading the nascent blog at Nature.com. A blog on web 2.0 technologies and science and science publishing. ANd then on June 8th Nature launched its Nature network Boston a " regional web site for the local scientific community". Its been only a few days still its launch ..and as a self professed web 2.0 geek I cannot wait for Nature to show us how web 2.0 technologies can help us network better in this city .

I have only started playing around with the site . The site is a very nice mish-mash of networking, blogging, discussions, groups events and all sorts of group driven things.

Since the site is so new..the first few days has seen very little activity but it will be nice to see how the site comes along. I have only started playing around with the site and found myself being the first to do a lot of things..It was very easy for me to start an "X-ray crystallography geek group" or join a "semantic web" interest group . Suffice to say the site makes it very easy to start interest groups and the idea as I see it is others join these groups and go on from there..

I did try to edit my profile and found a few of the edit options a bit buggy and not "editable" , but I am sure this a "beta" issue.

As with all things "community" , a lot depends on its constituents. Having Nature magazine host this effort will hopefully give it the legitimacy it needs among scientists to join the "social networking" bandwagon.

Science review American Idol style?

I first came across Nature magazines take on a new way to review papers at the business|bytes|genes|molecules blog. Since I am an academic scientist , I could not contain my enthusiasm for a new way to get my papers peer reviewed and was very happy to see Nature magazine taking the lead on this. On reading the FAQ associated with this trial peer review process. My enthusiasm was quickly dampened and my initial thought manifests itself as the title to this blog post.

But that said, I am big fan of Natures adoption of web 2.0 and the leadership role its taken with regards to its adoption in the scientific publishing world. This trial does seem like a sincere attempt at trying new things. There is however one big problem. Confidentiality . We all know how painful it is to arrive at that big result. Also typically, only the truly big breakthroughs end up being published in Nature. And though I have found myself complaining many times about how editors just don’t appreciate the novelty and “sexiness” of my scientific finding, I don’t think I would ever submit my paper to be “peer reviewed” online for the reason of confidentiality .Before something is published, sadly, and typically most findings are kept fairly under wraps. Pre-publication information is revealed only during meetings and posters and other such non peer-reviewed fora ( and that too only if your lab is not the super paranoid kind).

I would absolutley dread sharing my hard-earned research findings for all to see before it is published with the online community. And the key line here is “before it is published”. This novel attempt by Nature makes that almost imposible.

I would instead really like NIH, NSF and others funding agencies or possibly even Nature itself, to promote frameworks for electronic lab notebooks . The idea being, that once something is published, then all the data relevant to that publication is made online for peer and public perusal. Almost like the recent decision to make a lot of the clinical trial data public, a move to make publishing scientists put all their data out there will immediately reduce instances of fraud and make tax payer funded science more accountable. Also, now we will all know how typical that “typical data” is , post publication .

Importantly, many times , it is the failed experiments and “outliers” that don’t end up in papers that tell us a lot about the system and represent valuable information.

Supplementary material at “vanity” journals like Nature already run into 10s of pages. It will be great to have an NIH/NSF mandated fixed format lab notebook be made available online as well upon publication.

CRACking the scene :High-throughput screens identify mediators of calcium currents

The Endoplasmic reticulum serves as the intracellular store for Ca+2 which is one of the major mediators of a signal inside a cell. Typically a signal arrives and impinges on a protein receptor on the cell surface and this signal is conveyed into the cell by a cascade of events which culminate in the release of Ca+2 from the ER stores mostly into the cytoplasm. In many cells like the immune T-cells, this dumping of stored calcium leads to the opening of calcium release activated ion channels on the cell surface which then restore the Ca+2 levels inside the cell . these Calcium release activated calcium currents ( the so called CRAC currents) were observed in a variety of cell types but the exact molecular nature of these current carriers ( the wondrous ion channels) were unknown to science. Now three papers in Nature , science and PNAS respectively have all identified one of the genes that plays some role in mediating these CRAC currents.

The reason this caught my eye was two fold: one , I recently started working on ion channels and two , all three papers used their versions of high-throughput knockdown screens to arrive at the exact same gene. These studies all used whole genome RNAi in drosophila ( the fruit fly) cells to abolish function. These approaches involve disrupting gene function , gene by gene , one by one using small pieces of rNA thrown onto cells and then looking for a change. In this case a failure to replenish Calcium levels in the cytoplasm after the stores had already dumped. Thus genes involved in CRAC would kill CRAC currents when their RNA pieces are introduced onto the cells ( all done in a 384 well high-throughput format) and these cells would not recover after losing their endoplasmic reticulum calcium.

It is widely believed that these screens are highly plagued by error and irreproducibility owing to their complexity. A big problem is false negatives and an equally big problem is false positives. Although a lot of these caveats are well known and often brandished by the anti-systems biology brigade in established science. Studies such as these point to the maturing of these high-throughput screen approaches to identifying the function of genes. It is also heartening to note that all of these studies followed through with their 20-100 possible initial candidates with secondary assays also conducted in a high-throughput manner (like SNP genotyping analysis and high-throughput patch clamping)and arrived at a the identical gene. This gene is possibly an important candidate gene responsible for the CRAC currents.

There is no doubting the fact that this initial identification will lead to the identification of other partners involved in these CRAC currents which are of key importance in understanding the molecular basis of t-cell activation and several diseases that result from impairments in these CRAC currents.

refs: The Nature paper by Feske et. al., The science paper by Vig et. al., and the PNAS paper by Zhang et. al.