The last R newsletter (volume 7/1, April 2007) has a solution to a long standing problem. The new package mratios can deal now with ratios of means of normally distributed random variables and ratios of regression coefficients arise in a variety of ways. For two-sample problems, the package is capable of constructing confidence
intervals and performing the related tests when the group variances are assumed homogeneous or heterogeneous.
Here comes the reference to a first molecular system biology paper on asthma – something on my to-do list for 3 years. The authors constructed a biological interaction network using a database of curated molecular interactions. Continue reading Hubs and superhubs in asthma: low activity
seems to be the motto of many bloggers although it goes probably back to Pliny.
“Kein Tag ohne Präparat” (no day without taxidermy) was also the motto of Rudolf Virchow that I found last week in the Medizinhistorische Museum at Charité Berlin. More on this fascinating collection can be found at taz. No day without DNA is being the modern translation…
Science blogs usually refer to a scientific paper. To increase the visibility of science blogs, e.g. for a reverse lookup by search engines like “find all science blogs to a particular paper” it would be useful if science blogs would include a defined tag to which paper they relate. A http link will only partially work as single articles may be found at duplicate sites (journal or the publishers site or even through agencies like OVID and PUBMED CENTRAL). Using the DOI identifier is an alternative. To recognize any source document I therefore propose the following (unofficial) IANA scheme to be included somewhere in the body of your post
scienceblog:doi:10.1371/journal.pmed.0040072:
If there is no DOI available, I propose to use the link instead
scienceblog:http:www.thelancet.com/journals/lancet/article/PIIS014067360209654X: Please note that there should be an extra “:” at the end of the string; alternatively you may use a white space.
05.05.2007 Automatic DOI number extraction from blogs following this convention is now available at the Science Blog Finder page – just enter you rss feed address to get your blog indexed every 24 hours.
A paper (that I found only recently) summarizes the responsibility of authorship in the life sciences. Sharing publication- related data is a key element of the life sciences and there is concern that in practice materials are not always readily available to the research community. U-P-S-I-D-E stands for “uniform principles for sharing integral data and materials expeditiously”. The authors come from major U.S. universities and companies and have developed 10 recommendations that should be in the curriculum of every PhD program – go to the executive summary at www.plantphysiol.org/cgi/doi/10.1104/pp.900068
Not all biobanks will need daily access to samples. Here comes a cost effect alternative – storage in permafrost regions. ZEIT magazine has an article about “Mine 3“, a former coal mine in the Arctic. Already the first 10,000 samples have been stored there at -3,5 degree Celsius. BBC and Wikipedia have also information about Svalbard Global Seed Vault Continue reading Low cost biobanking
Allergy (or at least an associated trait) may have its roots somewhere in Africa – where helminth infections are frequent. A new Nature Immunol Review has an overview but I am quite disappointed. From the abstract Continue reading Wormy world
During my visit last week in Berlin, I found a remarkable letter, that corrects some misbeliefs about the allergy prevalence at that time. Here is my transcript – use Babelfish to translate it. Continue reading Allergy research in Germany 1935
Having identified earlier the largest problem in epidemiology I trying now to identify the second most relevant problem. Hopefully I can do it by just one sentence: Continue reading The second largest problem in epidemiology
1 MB is 1 Megabyte is 1,000,000 bytes and 1 Mb is 1 Megabase is 1,000,000 nucleotides. Although a new nature paper doesn’t make any fuss about it, there are 3,93 mutations / Mb in cancer tissues (in total they found 1,007 mutations by scanning 274 Mb from 210 cancer tissues). Continue reading 3,93 mutations / Mb
All4quotes and becontent have (mainly German) quotes. My all time favorites are
Aurelius Augustinus: Irren ist menschlich, aber aus Leidenschaft im Irrtum zu verharren, ist teuflisch.
August von Kotzebue: Menschen irren, aber nur große Menschen erkennen ihren Irrtum.
Christian Friedrich Hebbel: Die Menschheit läßt sich keinen Irrtum nehmen, der ihr nützt.
Friedrich von Schiller: Liegt der Irrtum nur erst, wie ein Grundstein, unter dem Boden, immer baut man darauf, nimmermehr kömmt er an den Tag.
Friedrich von Schiller: Hundertmal wer ich’s euch sagen und tausendmal: Irrtum ist Irrtum! Ob ihn der größte Mann, ob ihn der kleinste beging.
Georg Christoph Lichtenberg: Nur der Betrug entehrt, der Irrtum nie.
Continue reading Errare humanum est
A funny question answered in Allergy.
After adjustment for age, sex, parental predisposition and social status, the risk of hay fever was more than double in subjects who lived together with a partner having the same disease (odds ratio 2.4 […]). If subjects lived together with an affected partner, the risk of developing the disease increased with the time the partners lived together (1-11 years, OR 1; 12-23 years, OR 1.8; 24-35 years, OR 7.4; 36-54 years, OR 13.7).
Whatever that means – improved recall, shared doctor, non random mating, shared food, transmissible agent in descending order – it has been noticed already 40 years ago by Montgomery Smith and Lloyd Knowler in the Am Rev Resp Dis 1965; 92:16:
ATG16L1 is now confirmed as a candidate gene for Crohns disease:
Specifically, the LD structure and association mapping around the most associated SNP […] rs2241880 implicated a region on chromosome 2q37.1 containing a single gene known as ATG16 autophagyâ€"related 16-like 1 (ATG16L1) […] Logistic regression analyses conditional on A197T in the family-based samples indicated that this coding variant can fully explain the association signal to this locus; thus, we consider this to be the causal risk variant.
Excellent to have this replication, although the argument above cannot convince me that this already a causal variant.
ATG16L1 seems to be most abundant in CD4+ and CD8+ cells and knockdown of the gene will lead to loss of S. typhimurium autophagy. Does rs2241880 really induce a loss of function and how does it relate to TLR7 and NOD2/CARD15 signalling?
The authors of another Nature paper believe that there is a primary NF-kB signalling defect in the Toll-like receptor activation by intestinal bacteria – see also KEGG pathways.
It will be much easier now to ask the right questions.
