With all the M$ systems out there you know that it'll be a while before Linux gets larger numbers of crunchers and of course increased support. My linux box crashed after an update/install and just don't have the time to fix it since I have to study the GRUB to get it to work right. You can catch more Fly's with sugar and water that with vinegar. Stay with GC and include yourself with the Linux crunchers on BOINC.

For one of the admins or senior members:

In MY Grid: My Projects:
Available Projects:
The list contains just four active projects for BOINC, that I can select, as follows:

FightAIDS@Home
Genome Comparison
Help Defeat Cancer
Human Proteome Folding 2 (still in BOINC repair)

Not listed (not all-inclusive) but in discussions are:

Carbone
SIMAP
Decrypthon

What happpened ?
PaulT

Hello PTabolinsky,
The WCG is just running 4 projects currently. You can find the entire list of projects available for BOINC at http://boinc.berkeley.edu/projects.php
The list of current distributed computer projects (BOINC and not-BOINC) is at http://distributedcomputing.info/projects.html - a very important website. Look it over.

Lawrence

I know where you're coming from in the points thing. However, it's much a personal choise to consider points vs project selection. One does what one likes best with available resources.
For me I don't even look at the points. What I look at are the numbers of conpleted work units I am able to send out. If I could send you all my points to make you more happy (I'm not saying your unhappy here) just more happy - - I would gladly give them to your no matter what project you undertake. It just so happens that the GC seems like what is needed now ( to me) to make more sence out of the other crunching projects. Witness the recent news about "Until now it was believed the variation between people was due largely to differences in the sequences of the individual "letters" of the genome.
It now appears much of the variation is explained instead by people having multiple copies of some key genes that make up the human genome." [By Steve Connor] see also: http://news.nationalgeographic.com/news/2006/11/061122-human-genetics.html
I hope you don't let the points thing cloud your choises to much. I believe the Tech's are working the problem nut with all things it takes time and people hours. :>)
Best and cheers
PaulT


--
Edited to fix the quote tags - nelsoc

The points are not my primary concern. I have several criteria that I weigh when deciding which projects get my CPU time. My first consideration is the benefits the project will have for humanity. If I think the project will not produce benefits even it they achieve their research goals then I don't crunch that project even if I know I can get lots of points from it. However, when I see 2 projects that have equal merit as far as benefiting mankind then I have a tie situation. To break the tie I then consider my second criteria... the points I can get.

I have found that when a project has taken care of the points issue it is usually run by very bright and innovative people who track down bugs and fix problems quickly. The crunching at those projects is smooth and quick. Where they claim to be out of alpha/beta stage and into production but haven't taken time to solve the credit issues I've found there is invariably a myriad of issues and problems and it seems to me to be an indication of less talented people running the project or perhaps a project that is inadequately staffed. I can handle the kind of things that happens at those projects when I'm getting paid money but not when I am the one picking up the tab.



If we don't get a handle on global warming then the people who die from AIDS and cancer will be the lucky ones because when we lose huge tracts of farm lands and global famine follows. Loss of life and human misery will reach levels we have never witnessed in our lifetimes nor our grandparents and great grandparents during either world war. Cancer and AIDS could be cut in half if people used available preventative measures, stopped smoking, adopted a healthy diet instead of eating junk food 3 times a day 7 days each week, exercised regularly, etc.


--
This post has been edited to comply with forum rules - nelsoc

Dear Colleagues,

A main difference between our proposal (GenomeComparison) and the SIMAP project is the
algorithm employed for the similarity searches. It is widely recognized by
the scientific community that the Smith-Waterman (SW) algorithm is the
most sensitive, although time consuming. Also one of the authors of FASTA,
William Pearson, reluctantly recognizes this
(http://www.techfak.uni-bielefeld.de/bcd/Lectures/pearson3.html). In
the paper describing SIMAP for the scientific community, it is written:

Quote ...The optimal solution to generate the similarity matrix would be the
exhaustive application of the Smith–Waterman alignment algorithm (16) and
the subsequent storage of any significant scores. Although efficient
implementations (17) exist, the computational costs are beyond
feasibility. Thus, heuristic approaches like BLAST (1) or FASTA (2) are
used to speed up the search for biologically meaningful hits in a database
and they became the most intensively used tools in sequence analysis... end Quote.

(SIMAP: the similarity matrix of proteins. Thomas Rattei, Roland Arnold,
Patrick-Tischler, Dominik Lindner, Volker Stumpflen and H. Werner Mewes.
Nucleic Acids Research, 2006, Vol. 34, Database issue.)

To overcome some of these problems, in particular the CPU time demanded by
the Smith-Waterman algorithm, in the SIMAP project FASTA is used as a
first approach and then use Smith-Waterman to refine the results, by
excluding comparisons with low scores (FASTA results) before executing the
SW algorithm. Our approach employs directly the Smith-Waterman algorithm
ensuring the best detection from the very beginning of the pairwise
comparisons, therefore achieving the optimal solution mentioned by the
SIMAP authors in their paper. Even though there are several groups working in the development of more
sophisticated algorithms for more sensitive and fast similarity searches,
so far the Smith-Waterman algorithm is still the most sensitive.

There are some additional aspects to our approach: we use consistent parameters and cut-off values throughout the entire comparison procedure, thus assuring an even scoring. This will allow for additional comparative statistical analyses, for example in the study of genome evolution. Parameters are set to allow for the addition of data blocks while maintaining the scoring results completely compatible.

The current phase is only the first step in our Genome
Comparison project, since the next step may be the comparison of genomic
regions that have the potential to be coding, but are currently not
recognized as such, which may be due to the algorithms used for gene
prediction, among other factors.

best regards,

The GenomeComparison Team
Antonio Miranda
Marcos Catanho
Wim Degrave