Tuesday, November 30, 2004

Favicon + Forum

Added a new Favicon for those who bookmark the site... prettier that way, especially on the Firefox Bookmark Toolbar! Hope you like it :)

I'm also looking to implement a forum, but I don't know if enough traffic pass by yet (I don't want to host one of those 2-posts-by-the-webmaster forum, if you know what I mean). It would be bioIT oriented with a little biotech here and there... something to build a community around. Show your interest in the comment section!


2005 Life Science Insights Top Ten List

LifeScience Insights released their "Top Ten List" (registration required) of predictions about Life Science in 2005. They focus on the IT division of Life Science (computational biology, bioinformatics, bioIT). Overall the predictions are positive, especially for bioIT (see #4), but I can't say I agree with a part of their top 10 (points #2 and 7, in particular). The whole top 10 is available in the full post!

Note that they also have an overview of bioIT for 2004.

1. Technology growth later in the value chain
Technology spending will increase relatively more in downstream processes, such as clinical trial management systems, electronic data capture, and electronic lab notebooks.

2. Disaggregation happens
Big pharma IT divisions are moving away from the “mother ship” model. “Individual business units are making more and more IT decisions,” said Golden. Any vendor with a solution can at least “get their foot in the door.”

3. Innovation will happen - somewhere else
R&D will increasingly become a target for economies, as pharmas look to outsource discovery programs to China, India, and Singapore. “Can we outsource medicinal chemistry and target discovery?” will be a key question in Golden’s mind. There will also be growing emphasis on capturing external innovation, for example in the form of literature mining.

4. The line between IT and drug discovery . . . will continue to blur
“Computational biology will go away,” said Golden, “because most biology will be computational.” He said this item could have been subtitled “Genomics doesn’t matter…” As aging pharma veterans retire, there will be a warmer reception for in silico prediction tools, although Golden said he still believes drug discovery is more an art than a science.

5. Thus, the rise of the Pharma CIO
This position will have a larger seat at the R&D table. While CIOs managing non-core businesses will be spending more time in Asia, funding sources will increasingly come from both corporate IT and the business unit itself.

6. Data standards will drive a fragmented industry
Although not required, a growing number of standards including SDTM (study data tabulation model), SAFE (secure access for everyone), and SEBIX (secure electronic biopharmaceutical information exchange) for submitting and managing clinical data will become “de rigueur.”

7. The curse of Eliot Spitzer
IT organizations are scrambling to comply with a growing list of regulatory mandates and meet deadlines. “Pharma lawsuits by states’ attorney generals will drive a new market for compliance software, and make Sarbanes-Oxley look like a picnic,” said Golden. Look for biopharmas to hire a CCO - Chief Compliance Officer.

8. Pharmacovigilance
The buzzword for the next few years, thanks to the Vioxx withdrawal and suspicion over other blockbuster drugs. “Pharmacovigilance is a problem of data integration,” said Golden, who offered a broad definition that included toxicogenomics, adverse event reporting, and regulatory compliance. “The FDA will want it; regulators will demand it; pharmas, biotechs, and the NIH will struggle to build it,” he said. The first vendor to offer a comprehensive solution, probably within the next 18 months, will own the space.

9. Lost on the NIH Roadmap
The 2002 NIH Roadmap, proposed by director Elias Zerhouni, could lead to a dramatic, but welcome, reduction in the number of NIH institutes, with other government agencies looking to fill the void. Golden sees a growing role for venture philanthropists such as the Bill and Melinda Gates Foundation.

10. Venture capital is back. We’ve missed you
“The pendulum is slowly swinging back to tools and platforms,” Golden said. He asked why VC firms tend only to look at MIT and Stanford? “I’d pitch a tent in Madison, Wisconsin, and probably make a lot of money,” said Golden. The successful $3-billion California embryonic stem cell initiative could bolster Series A financing.

Thanks @ Bio-IT World.


Sunday, November 28, 2004

Therapeutic siRNAs

Interesting article on Biomed Central. siRNA are getting big in the research world; they allow to easily (and most of the time, specifically) knockdown expression of a gene by 50-95%. They are extremely usefull to study protein functions, but they also show lots of promise in therapeutic applications.

Imagine being able to knockdown expression of a viral or bacterial gene long enough for the immune system to get rid of it. Or knockdown expression of a gene linked to drug resistance in cancer cells. Until now, a (potential) problem undermined research in this area. siRNA are double-stranded RNA molecules; receptors (Toll-like receptors) on some cells recognize dsRNA as foreign (double-stranded RNA aren`t expected outside cells) and trigger an interferon-mediated immune response. Scientists feared that siRNA would trigger this response; when injected on a large scale in a living being, the intensity of it would be too strong and cause damage (or death).

Scientists being what they are (curious), they had to try. They injected siRNAs designed against Luciferase (a firefly protein hydrolyzing luciferine, producing light) along with the plasmid encoding it (with appropriate negative and positive controls). Then they mesured IFN-gamma levels (elevated in case of a TLR-mediated immune response). Not only did the siRNA reduce levels of luciferase appreciably, but a total absence of IFN-gamma response was observed. It seem that the length (20-22 bases) of siRNA could explain the phenomenon. I don't need to tell you that these are excellent news for future siRNA-mediated therapy.

Read the paper on Nature Biotech.


Saturday, November 27, 2004

Canada Biotech Sector Growth

An article on Reuters nicely summarize the state of the Canadian Biotech sector. Promise of growth (or should we say recovery) of the Toronto Stock Exchange health sector rely on 30 or so Biotechs successfully delivering products they are developing.

Among them :

"High on the list is Cardiome Pharma Corp. (COM.TO: Quote, Profile, Research) , which is due to report data from the first of its pivotal studies of its RSD1235 treatment for atrial fibrillation, a coronary disorder that can result in a stroke.

Inex Pharmaceuticals Inc. (IEX.TO: Quote, Profile, Research) expects to find out on Jan. 15 whether the U.S. Food and Drug Administration will approve its lead product, Marqibo, which treats non-Hodgkins lymphoma, a group of malignant diseases that develops from the lymphatic system."

Anormed Inc. (AOM.TO: Quote, Profile, Research) , fresh from receiving an $18 million payment for the kidney medicine it patented, hopes to start pivotal trials in the new year for AMD3100, a drug designed to stimulate stem cell production.

Note that the whole Canadian Biotech sector represent bout 4% of the global marketshare, compared to 77 percent for the USA. More good news for research folks :

The Canadian government has been forecasting growth for biotechnology. It poured C$695 million ($585 million) into the industry in the year ended on March 31, 2003, a 24.8 percent increase from the previous year and a 60.8 percent jump from fiscal 2000 and 2001.

Ottawa also plans to double its investment in research and development, including biotech.

Back!

Finally home... the meeting was kinda nice, except for some minor points. First, no posters, only oral presentations... so our chance of interaction with members of other teams was minimal. Second, my presentation crashed! Note to self : don't encode fancy movies in DivX when not presenting with my computer... their laptop didn't have the proper codecs and displayed blank placeholders :( Next time be sure I'll be bringing my own laptop along... It was informative to see the projects of other AIDS Quebec-based teams, at least, and we sure had lots of fun (rented PT cruiser + crazy students, guess the rest :)).


Thursday, November 25, 2004

Away!

Just a little post to inform you that I'll be away for the next 2 days. I'll be attending (and doing a little presentation) a provincial AIDS research-oriented meeting. I'll be back posting news on saturday!


Tuesday, November 23, 2004

Bioinformatics 2.0 for MathLab

The MathWorks just released version 2.0 of their MATLAB Bioinformatics ToolBox. Its a set of bioinformatics-related functions allowing to do various tasks, ranging from microarray data normalization, visualization and analysis, protein, DNA and RNA sequence analysis, phylogeny, data retrieval from the web databases, etc. Its similar to Accelrys GCG Wisconsin package, but a few key differences should be noted.

First of all, functions are implemented in MATLAB language, are open source and customizable. The MATLAB environment also allow you to develop (and share) your own algorithms. It offers advanced machine (vector-based) learning options, mass spectrometry and microarray analysis. The full list of functions can be found here. Note that it isn't free (prices start from $1000 US), but a demo is available. It is available for Windows, UNIX/Linux and Mac platforms. Since some people getting into bioinformatics come from an engineer/math background, the opportunity to develop and work in a MATLAB environment is interesting. Excerpt from the press release :

New Toolbox Features Mass-Spectrometry and Statistical Inference and
Prediction Capabilities That Enable Faster and More Customized Analysis of
Data

Addressing the needs of computational biologists and bioinformatists, The MathWorks today announced the availability of the Bioinformatics Toolbox 2.0 for MATLAB(R). Scientists and researchers can now perform mass-spectrometry data analysis, perform statistical inference and prediction, view graphs, and conduct enhanced genomic and proteomic sequence analysis. [...]

The Bioinformatics Toolbox 2.0 offers computational biologists and other research scientists an open and extensible environment. Most functions are implemented in the open MATLAB language, enabling users to customize the algorithms or develop their own. The new mass-spectrometry data analysis feature is specifically designed for pre-processing data, including baseline correction, smoothing, alignment, and re-sampling. [...]

The Bioinformatics Toolbox 2.0 also builds on the classification and statistical inference and prediction tools in the Statistics Toolbox by providing several new classification functions and tools for identification of discriminating features, and visualization of complex data is enhanced with new graph-viewing functions and manipulation tools that display interaction maps, hierarchy plots, and pathways. Additional features of the Bioinformatics Toolbox 2.0 provide access to specialized visualization tools, ranging from sequence alignments and microarray principle component plots to building and interactively viewing and manipulating phylogenetic trees. [...]

In an effort to bring computational biologists, bioinformatists, and other technical professionals the power and performance they demand in their applications, The MathWorks recently released the Distributed Computing Toolbox, which enables users to execute MATLAB algorithms in a cluster of computers. [...]


Saturday, November 20, 2004

Alternative Splicing Microarrays Incoming

Agilent and ExonHit Therapeutics announced a research partnership aiming to allow detection of gene splicing variants. Alternative splicing allow for different proteins to be produced from an unique gene. Its an additional layer of biological complexity that we're only beginning to understand.

Sadly, current microarrays can't differentiate between spliced transcripts; they usually only differ by the presence/absence of a coding segment (exon), the rest is 100% identical. Combining Agilent microarray platform (60-mers cDNA-based / one probe per gene, sadly) with ExonHit DATAS technology promise to eliminate this shortcoming.

Excerpt from the press release : "Agilent Technologies Inc. (NYSE: A) and ExonHit Therapeutics, a private drug discovery company, today announced a research collaboration to combine Agilent's microarray platform and ExonHit's alternative RNA splicing technologies and expertise. This collaboration explores the development of a microarray-based solution that will enable scientists to properly monitor the expression of splice variants.

Splice variants are variable sequences of RNA produced from the same gene in DNA, resulting in the creation of different proteins potentially affecting cellular regulation. Scientists developing therapeutics are increasingly interested in this emerging field as the expression of splice variants can provide novel targets, may indicate disease states, and can be altered by exposure to drugs and toxins.

Agilent and ExonHit are working together to optimize microarray design, reagent protocols and data analysis methods for splice variant studies. As a pioneer in alternative RNA splicing, ExonHit realized that the proper characterization of splice variant expression required dedicated profiling platforms. The company has received notice of the allowance of its patent, which broadly claims nucleic acid arrays that enable the detection of alternative RNA splicing events via either intron or exon and splice junction-specific probes. Agilent is a leader in the gene expression field, providing researchers highly flexible and sensitive printed microarrays.

Initial results from an experimental splicing array of G-protein coupled receptors, designed by ExonHit and produced by Agilent pursuant to the collaboration, were presented at Splicing 2004, an annual symposium on alternative RNA splicing, by Richard Einstein, vice president of R&D North America at ExonHit. The array detected multiple isoforms of several genes, and showed good reproducibility and specificity. The companies are expected to work with early test sites to generate additional experimental results."

Personally, I think that Affymetrix technology is superior to solve the alternative transcript quantification problem; I sure hope that they give it a shot. With more probes (typically 12x2/gene) than Agilent microarrays, even if they're smaller (12-mers vs 60-mers) would allow for a more precise quantification by tiling known exons... I'll wait for the actual product from the Agilent / Exonhit partnership and their implementation to judge tough.


Wednesday, November 17, 2004

Influenza Genome Sequencing Project

Along the lines of the Human Genome Project, and related to our previous story about a potentially imminent pandemic of Influenza, The National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health (NIH), announced yesterday their Influenza Genome Project.

Considering that Influenza (Type A) genome is a measly 13.6 kb of single stranded RNA (divided in 8 molecules), compared to the 3 billions basepair for the Human genome, it seems like no big deal; a quick search in Entrez reveal that we already have the sequence for 14 Influenza genomes. With this project, the researchers intend to sequence thousands of strains of Influenza (Avian and human viruses). They collaborate with institutions holding collections of viruses accumulated during several decades; among them, the St. Jude Children's Research Hospital, where the research will take place. With this project, they hope to improve our knowledge of the virus. Multiple alignment studies correlated with infectious potential may help to understand which mutations are more dangerous than others. More importantly, the database will be a great tool for vaccine development.

Story coverage on MedicalNewsToday


Tuesday, November 16, 2004

NASA + Active Genome Mapping

NASA isn't just about rockets and Mars-crawling robots... The NASA Ames Research Facility just published a very interesting genomics study. They designed High-density Microarrays, based on the recently "finished" Human Genome Project, using their Maskless Array Synthesizer (MAS) system (manufacture of Affymetrix arrays use mask-based lithography). 134 arrays and 52 millions unique probes (spaced on average 46 nucleotides apart, spanning 1.5 Gb of non-repetitive sense and antisense genomic DNA) later, they had a "map" of the entire genome. These tiling arrays were then used to determine experimentally and with great precision the regions expressed as RNA (in liver tissue), corresponding to genes (and all their splicing variants) and microRNAs (or something else yet to be identified).

Transcription was detected from 64%, 57% and 35% of known genes from RefSeq, Ensembl, and predicted genes by GenScan. Not every gene is expressed in every tissue (~50% of known genes seem about right, from my experience with Affy arrays and CD4+ T cells); the low percentage from GenScan prediction may indicate a lot of false positives by the gene prediction algorithm. They also identified 10,595 transcriptionally active regions (TARs) not present in databases. They confirmed 90 (out of 96 tested) of these by RT-PCR. By BLAST analysis, they found that many of their novel TARs (24%) had significant homology with mouse Genome sequences. 14% of the novel TARs, when translated to protein in all 6 reading frames, showed significant homology with mouse proteins. Their "expression map" is quite an achievement... you can read the original paper here.

Recent NASA Ames Genome Research Facility publications
Coverage by SpaceFlight Now


Sunday, November 14, 2004

Incoming Influenza Pandemic?

According to the World Health Organization, an Influenza Pandemic is incoming. This is unrelated to the fact that there is a shortage of influenza vaccine in the US (The vaccine in short supply is targeted at a different strain than the feared pandemic-causing one). Vaccine-producing companies throughout the world scramble to design a solution against the strain H5N1 ("H" stands for Hemagluttinin and "N" for Neuraminidase, two viral proteins), who could migrate from birds to humans with devastating effect. Should we be scared?

The answer's both yes and no. Yes because it did happen in the past, and the consequences were catastrophic on a worldwide basis. The Spanish Flu of 1918 killed 25-40 millions people in 18 months. 20% of the world total population got infected to some degree; it affected everyone, not just the traditional groups at risk (old and very young people). If it would happen again, with modern rapid-travel (airplanes, etc), chances are it would be worse, and faster. No medication is efficient at stopping the virus itself, only the symptoms (fever, headaches, sore throat) are targeted by drugs to somewhat relieve infected individuals. An efficient vaccine is still the best bet against flu. Everyone should get vaccinated each year; even if you're not part of a risk group, it helps diminish the spread and mutation rate of the virus.

Links of interest :

US weekly Influenza Report
Killer Flu - Secrets of the Dead
The American Experience


New post style!

Long posts cluttered the main page, so I implemented a summary/link combo for them. You can always click on the post title or the "Read more!" link to get the whole story. Plenty of advantages... the main page is less cluttered, so you can see past posts more easily. Ads are more targeted everywhere, so they are more relevant to your interests. Speaking of ads, I finally fixed the problem with Adsense! thescientistblog.blogspot.com got Public Service Ads (non-paying, untargeted, which I replaced with Amazon Bioinformatics Books) while www.thescientistblog.blogspot.com got targeted ads about bioinformatics, AIDS, HIV and biotech in general. So update your bookmarks according to your preference!


Saturday, November 13, 2004

Genomatix 3.2 - Chicken and Rice Genomes

This week Genomatix released version 3.2 of their excellent promoter-analysis software, GenomatixSuite. This follow their last month release of the new and improved GEMS Launcher 4.0. According to the press release, new features / improvements include...

ElDorado / Gene2Promoter
* New genomes:
o Gallus gallus (NCBI build 1 v1)
o Oryza sativa (TIGR release 2)
* Genome updates:
o Human Genome updated to NCBI build 35.
o Mouse Genome updated to NCBI build 33.
* Display of alternative transcripts of a gene in the ElDorado result.
* MatInspector: Export of statistics and common TFs to Excel format.

So now you can do promoter / regulation analysis on the Chicken and Rice Genomes (I'm sure you were all waiting for this, badly, no?). Recent builds of more important genomes (Mouse & Human) fix some errors of previous builds (misallocated contigs, etc).

More importantly, spliced transcripts are viewable from ElDoraro (their Genome Annotation portal), which is very important when doing microarray analysis. Most microarrays don't differentiate between alternative transcripts, and the gene ID often refer to only one of them, so it's good to know that they exist. Most of the time, spliced forms have different functions than the major protein (they can even be implicated in competition / inhibition of the pathway they're part of : no functional domain but can still bind its partners, for example).

I even encountered a case of an upregulated gene, but the most documented protein was (for a long time) the spliced form, which inhibit protein X functions. The full-length protein was discovered only recently, and it turns out that it stimulates protein X. Go figure. No wonder it was the only thing going against all my previous analyses conclusions... now everything is fine. Keep an eye out for these special cases... I think they'll get more frequent as I'm sure we underestimate the role of splicing in a cell.



Wednesday, November 10, 2004

Trinity Biotech's HIV test gets FDA clearance

Trinity Biotech, an Ireland-based company, issued today a press release announcing that they just got FDA approval for being used in doctors' offices and clinics. Previously, it could be used in the hospital environment only. Why is this important?

Trinity Biotech’s Uni-Gold™ Recombigen® HIV test is approved by the FDA for the detection of antibodies to HIV in human serum, plasma, venous and finger stick whole blood and is the only Rapid HIV Test to be approved for use with all four sample types. The product is easy to use, requires one-step and produces a result within ten minutes.
So basically, its the quickest AND best test on the market; and now, it's even more available. Testing is important; in developped countries, sexual tranmission of HIV-1 occurs principaly with a partner unaware of (his/her) infected condition.


Toucan 2.0

I received a mail informing me that TOUCAN, a regulatory sequence analysis package, just released a new version. More importantly, they've gone Open Source, which is good. I tried version 1.0 last year and had mixed feelings, notably about the Java interface. Version 2.0 brings lots of improvements on this side (the interface is still ugly, but its more functional), among other things. The feature set is impressive for anyone doing promoter regulation studies (automatic promoter retrieval, TRANSFAC database querying, overrepresentation analysis, etc). I pasted the announcement of the release, which include a changelog.

Dear TOUCAN user, Dear MotifSampler user,

We are pleased to announce the second release of the TOUCAN software for cis-regulatory sequence analysis, and the 3.1 release of the MotifSampler algorithm. All information about TOUCAN can be found here:
http://www.esat.kuleuven.ac.be/~saerts/software/toucan.php
All information about MotifSampler can be found here:
http://www.esat.kuleuven.ac.be/~thijs/Work/MotifSampler.html

1. Launching TOUCAN 2
Because you have entered your email adress before, you can now simply launch TOUCAN 2 by clicking here: http://www.esat.kuleuven.ac.be/~saerts/software/toucan.jnlp Or you can use your TOUCAN shortcut to version 1 (created by Java Web Start) if it is still on your desktop. This will automatically download the latest version. Note: this will only work if you have JAVA 2 and Java Web Start installed, see the web site for further information.

2. New Features in TOUCAN 2
Some of the new features in Release 2, since version 1.28.0, include:
* ModuleScanner service: find the best N putative target genes of a cis-regulatory module in the genome (max_bp_length + a combination of PWMs as input). Returns top N sequences in a new window. This method is useful to validate a newly found module of the ModuleSearcher service.
* Retrieve all Ensembl-available orthologous sequences for a set of genes automatically
* New services added for sequence alignment: align 2 sequences with AVID, LAGAN (both global alignment) or BLASTZ (local alignment), and compare the results.
* Align all orthologous pairs of sequences automatically in one job, using AVID, LAGAN, or BLASTZ. Ideally suited for sets of co-regulated genes. Then detect over-represented motifs in the conserved non-coding sequences.
* Multiple alignment of an orthologous family by MLAGAN
* Menu items have changed, for example all motif (transcription factor binding sites) related tools are grouped.
* Zooming is improved: easy zooming by pressing Ctrl+ or Ctrl-
* Faster sequence retrieval from Ensembl by omitting the retrieval of external references
* After sequence retrieval from Ensembl, the "RevCompl Negatives" function reverse complements automatically all sequences that have the CDS on the negative strand.
* Open/save files: remember the previously chosen directory

Some of the new features that appeared gradually between the original publication in NAR and version 1.28.0

* FootPrinter 2.0 web service
* ModuleSearcher web service, A* and GA version
* Run MotifScanner on sublist sequences
* JASPAR position weight matrices available
* Automatic selection of features for all sequences

3. Open Source

TOUCAN 2 has become open source software. By releasing the source code we hope that other people will contribute to the TOUCAN project, so that TOUCAN can provide
- web services to all the latest algorithms (e.g., for multiple alignments, phylogenetic footprinting, binding site prediction, cis-module prediction, etc.)
- better functionalities and visualizations
- better integration with existing software or databases (e.g., genome browsers, DAS servers, etc.)
- general robustness and ease-of-use, also for biologists
If you would like to contribute to the project, please send me an email. Communications around the project can go via the mailing list (join here:
http://listserv.cc.kuleuven.ac.be/cgi-bin/wa?SUBED1=toucan&A=1)

4. Service Mirrors

We have prepared an installation package for the web services, so that they can be made available from other locations. We expect that this will increase the services uptime significantly. If you would also like to mirror the services, please send us an e-mail and we will forward you the installation instructions.

All suggestions and comments are welcome and we hope you will find your cis-acting regions using TOUCAN !

Sincerely Yours,
The TOUCAN Team.
The MotifSampler Team.

Monday, November 08, 2004

UK PhD opportunity

My friend Daniel over at Bioplanet asked me to publish a (very good, in my opinion) Ph.D. opportunity. Genomics and Bibliomics are currently very hot subfields of Bioinformatics. Here's the ad...

PhD Studentship at Cardiff University, UK

The PhD title is "Application of improved automated text mining to cancer transcriptome datasets" and centres on the development of improved mechanisms to mine gene-associated metadata (PubMed, OMIM, etc) associated with groups of genes identified through mass-parallel 'omic techniques. It will use real datasets that are being produced on the Affymetrix gene expression platform through our CRUK Programme Grant, but the techniques would be generally applicable to other techniques (e.g. proteomics, Taqman arrays, etc).

The student would join a small but active bioinformatics team and the other members of the CRUK Programme in our new Henry Wellcome Building at the School of Medicine. It would particularly suit a student with existing mathematical, statistical or computer science skills who wished to broaden their experience into the biomedical field.

Further details of the studentship can be found at:

http://science.cancerresearchuk.org/gapp/grantapplications/tcdb/tcd_phd

Supervisor: Professor David Kipling
Contact: KiplingD@cardiff.ac.uk


Sunday, November 07, 2004

Intelligent Drug Design

Syrrx, a drug design company, just released the 3D structure for their main target, human 11(beta)-hydroxysteroid dehydrogenase type-1 (HSD-1). Inhibitors of this enzyme have potential in the treatment of type II diabetes. Why is this interesting? Consider the drug design pipeline :

Traditionally, the process to identify new compounds with interesting medical properties was based on luck; throw a bunch of randomly synthetized molecules, or earth/bio/plant samples, on a column containing your target of interest, and retain those who bind. Then check for medicinal properties, do some animal testing, check for dangerous degradation products by the liver, various pharmacokinetics parameters, do some modifications to improve it all... 10 years and 800 millions US$ later, you have a new drug!

When I first learned about the 'random' drug design approach, I remember being quite disappointed ("Is that the best we can do?"). The new trend is intelligent drug design. You determine the structure of the enzyme / receptor you're targeting, based on previous studies. Then, you design a compound that will bind to the active site (or prevent dimerization, etc.) with chemical groups that will assure good systemic distribution... much more intelligent than the old process, and far less random and costly! As you can guess, Bioinformatics play a crucial role in intelligent drug design... so it's good to see it become reality.


Zlabs Gene Regulation Tools

Just came across Zlabs Gene Regulation Tools, a site regrouping links useful in gene regulation analysis. Promoter databases / extractor, overrepresented motifs finders, promoter prediction, microarray analysis, phylogenic footprinting, it have it all. Great portal, be sure to check it if you're in this kind of stuff! I'm adding it to the "Tools" section right now.

As a side note, I entered the BioBanner program to replace Google Adsense ads when they go non-targeted (The "Save Gorillas" stuff, great cause, but no ad revenue from those, sadly). Its an banner-exchange initiative put up by the folks at bioinformatics.org for biotech related website. Let's hope that it'll increase incoming traffic up here... and let's hope the Google bot recrawl the page soon :( No ad revenue = Sad Panda.


Friday, November 05, 2004

NIH funds new bioinformatics resources

There's been a lot of talk about the (over)hype of bioinformatics as "The Way of The Future" and the apparent lack of concentrated effort / funding to justify the media buzz... creating lots of uncertainity / discomfort in the field. Well, good news are on the way...

As reported by MedicalNewsToday :

"The National Institutes of Health (NIH) has awarded two collaborative contracts, totaling $46 million [...] the funds will be allocated to UT Southwestern Medical Center at Dallas to create two bioinformatics centers.

The contracts will create separate bioinformatics resources - the BioHealthBase Bioinformatics Resource Center for Biodefense and Emerging/Re-emerging Infectious Diseases, and the Immunology Database and Analysis Portal (ImmPort). Each will include databases and software that researchers can use to improve drug discovery and vaccine development."

Biodefense and infectious disease are getting lots of funding lately (think bioterrorism), which is a good thing. They'll put the focus on tuberculosis and Influenza (common cold virus) apparently. The "BioHealthBase" part will concentrate on promoter analysis (at least, this is what I understand from the scientifically dumbed down article) while the "ImmPort" part will try to "integrate data from different kinds of biomedical research, such as clinical trials, animal modeling and studies using cultured cells", leading to a better global understanding of diseases.


Small redesign

Redesigned the site banner... the virus was too fuzzy for my taste at a resolution that low. Removed the texture on the text too, as it was too distracting / not enough crisp.

Sorry!

Sorry for that last triple post, it seems Blogger got unresponsive last night (it just came back up) and did all kind of crazy things. Everything is fixed now, so we can get back (soon, got infections to do!) publishing news :)


VIBE for Mac OS X

VIBE (Visual Integrated Bioinformatics Environment) from Incogen just got released for Mac OS X. Previously, it was available for Unix and Windows platform only. Apple being popular within the scientific world (for some obscure reason I can't quite explain) probably justified the move. In case you wonder :

"The INCOGEN Visual Integrated Bioinformatics Environment (VIBE) is a state-of-the-art, drag-and-drop analysis workflow management environment."
If you wanna see VIBE in action, a tutorial flash video is available.

Basically, its used to build a data analysis pipelines with little to no effort. A powerful drag/drop interface allow you to build graphically pipelines, with conditions / parallel analyses. Its a lot better than scripting command line functions. Of course, only labs doing high throughput sequence analysis (lots of BLAST, FASTA, Gene prediction, etc) can justify buying this software (it isn't exactly cheap), but its interesting to see that Mac still has developers attention in the bioinformatics world.


Thursday, November 04, 2004

Meet... ChiliBot!

As you may know, I have a special interest in Bibliomics (science of scientific litterature). When you're analysing microarray results, you're confronted with a big list of modulated genes, and you want to extract the biological signification of it. The main source of information is Pubmed by far, but genes (and proteins) often have 5-10 synonyms... multiply this by 200 modulated genes, calculate the number of pairs of genes to look at (to see if known interactions exist), and you got a big problem.

Fortunately, bioinformatics programs to ease the pain just start to show up. You may already know Bibliosphere, by Genomatix. Quite excellent, but has its limitations (the number of analysis / month is limited). Enters Chilibot (OK, lousy name). Noticed the article describing its conception in BMC Bioinformatics last week, and had to try it. And let me say, its wonderful. Not perfect mind you... but close. You give it a list of genes AND/OR biological concepts (apoptosis, T-cell, HIV-1)... it finds automatically ALL synonyms and perform pairwise searches on Pubmed, then output the result in a very nice graph. Free. On the Web. Try it. Now!


Tuesday, November 02, 2004

Elections Day, Bioinformatics and Immigration

Elections Day in the USA... MY GOD its a complicated process! Up here in Canada we vote for our deputy... and thats it. They vote for their president, senator, governor, house representative, a few state-dependant bills... by punch card, electronic voting booth... Honestly, I'm looking at the various maps on CNN.com, and I wouldn't have a clue if I had to explain the whole system. Canada 1, USA 0!

Important stuff aside, it seems that the province of Quebec wants to develop its (economic) relations with India, especially in the bioinformatics area. The article state an obvious barrier : Indians can speak English (not fluently, but oh well; personal judgment, from most indians posters in bioinformatics boards), but can they speak French fluently? Immigration laws in Quebec favorise (by far) french speakers with a good skill set; for this reason, most Indians coming in Canada choose Toronto or Vancouver. Even if they wanted to come here, quite frankly, I don't know what to think about it. Wouldn't be more productive to exploit the pool of non-working IT workers (coming from the dotcom era) up here than to go fetch some across the globe? Globalization they say... I'm not sure its a good thing in the long term.


Monday, November 01, 2004

Smallpox secrets partially unveiled

The New York Jewish Times reports that a NIH-funded study, published in PNAS last month, unveiled some smallpox secrets using microarrays. As you may know, smallpox is a potentially deadly virus who have been eradicated thanks to an extensive vaccination campaign. The only problem is that my generation (young adults) wasn't vaccinated and authorities fear that smallpox could be intentionally released by terrorists organization as a biological weapon (it would cause more fear than harm, but anyway). The only time it has successfully been used as a weapon of War was in 1754-67, by British forces who gave infected blankets to native Americans; some tribes lost 50% of their population.

The study used microarrays (the cDNA kind) to characterize changes in PBMCs (Peripheral blood mononuclear cells - T & B cells, macrophages, monocytes, etc) from 22 sinomolgus macaques (a study published in the same issue of PNAS shows that they can be infected with smallpox and develop a disease similar to human smallpox - previously, no animal model was available). From the abstract (will have access to the full text tomorrow) :

"Of particular interest were features that appear to represent an IFN response, cell proliferation, immunoglobulin gene expression, viral dose-dependent gene expression patterns, and viral modulation of the host immune response. The virtual absence of a tumor necrosis factor alpha/NF-kappaB-activated transcriptional program in the face of an overwhelming systemic infection suggests that variola gene products may ablate this response. These results provide a detailed picture of the host transcriptional response during smallpox infection, and may help guide the development of diagnostic, therapeutic, and prophylactic strategies."

Understanding the virus, even if its extinct, is good in case we have to face it again. According to www.TerrorismAnswers.com :

"Which countries have smallpox stockpiles?
Officially, smallpox remains in only two places in the world: the CDC in Atlanta, Georgia, and a repository in Russia. In 1971, the Soviet Union tested a biological weapon containing smallpox; it killed three people and had to be contained by an extensive vaccination campaign. Recent reports indicate that in the early 1980s, Russia grew large quantities of smallpox and successfully adapted it for use in bombs and missiles."

"Are the Russian smallpox stockpiles safe?
Since many laboratories in Russia are now financially strapped and are downsizing, the United States is worried that underpaid or unemployed Russian scientists and researchers might sell, or have sold, bioweapons expertise or equipment to terrorists or rogue states."

But don't panic just yet : the virus is difficult to produce, cultivate, almost impossible to dissipate on a large scale (stay infectious for ~24 hours, less under the sun). Vaccination (stockpiled) is available, symptoms are difficult to miss (a BIG rash).. etc. It would be effective the way anthrax was : create a massive paranoia, paralyze the targeted country, and cause very few deaths. A terrorist weapon; ineffective, but with a big scare factor.