evolutionary-research site news

News

Recovery from hard disk problems(10 Apr 2006)

Due to a hard disk failure evolution@home lost all results emails submitted between the 21st and the 31st March 2006. If you still have your submissions from that time, please send them again.

What happened? Without going into too much details: A new, bigger hard disk was employed to accomodate growing results-submissions and it worked fine for about 2 weeks. After leaving the computer with some long computations, I returned from a conference only to find it crashed for unknown reasons. After rebooting I started to check email (downloading most results from the server) and in the process the hard-drive quickly deteriorated (another crash > unable to reboot > access to some lost data still possible by rescue tools > after some rescue attempts not even that was possible any more). So I had to play in my total backup from the 21st of March and build again from there. Unfortunately several serious attempts to recover submitted evolution@home results were unsuccessful.

So, if you still have some of those results (21-31.3.2006 only), then please send them again. There is no need to send other results twice, as duplicate submissions are generally discarded.

This incident raised the question as to what has to be done to burn in a new hard drive before it can be trusted? If you have more experiences than can be found by a simple websearch, then please let evolution@home benefit from your expertiese (contact). I already know that

Ideally I should use RAIDs (problem: they are currently out of the financial reach for evolution@home, so I am seeking an inexpensive short-term solution),
Most electronical cirquits fail either quite quickly or after a long time (bathtub curve). What I am looking for is a mechanism to catch the quickly failing devices without ever trusting them with any data (with or without a RAID).

Ideally I need an inexpensive software tool that could do that reliably. Do you know one? Let me hear.

Project 3 enters the long-term computing phase (30 Jan 2006)

More than 3.5 years of CPU time have been submitted since the start of Project 3. For the moment the simple tasks have been computed with enough accuracy. Now computation will have to focus on the more complex tasks that will take several CPU-decades to resolve.

Project 3 looks at the effects of Muller's ratchet if combined with another ratchet process that has different mutational parameters. A thorough analysis requires the computation of a very large number of parameter combinations (See here for statistics). Since most of these processes will be operating in fairly large populations in nature, their computational modelling is again a challenge worthy of evolution@home.

Practical tips for the submission of results (1 Nov 2005)

The submission of results by email as it has been described earlier works fine for most results that have been received. However, occasionally strange combinations of email-programs and servers lead to problems that change results in such a way that they cannot be used for further analysis. Zipping of results avoids this.

While most results submissions work fine as they are, occasional problems have rendered a small fraction submitted results unusable for further analysis. The reason is that some email-clients and some email-servers modify the content of the emails if they think e.g. that lines are too long or line-endings are of the wrong type for the operating system they run on. Unfortunately the results-format of Simulator005 results is quite sensitive to these issues. Development of a more robust format is planned as part of the transition to full automation. In the mean time the following practical tips might help to make results-submission easier:

Zip results. Experience shows that the email-transmission procedure handles zip-files with more care than actual text files, since it does not try to be 'clever' when handling zip files. You may use any zip program of your choice to compress results with the zip algorithm (Probably you have some zip tool already as part of your operating system; if not search Google for "free zip tool").
If the resulting zip file cannot be opened for some obscure reason, then you will receive an email asking you to resubmit these results (preferrably using another zip programm). If you get this kind of feedback, then you will most probably get it around the time of the next high scores update. Therefore it makes sense to store your results for some time.
Store results. In the rare case of email submission problems, your results cannot be used. To avoid loss, you need to store them on your hard-disk and resubmit them, if asked. You can certainly delete all results that have already shown up in the high scores.
Remove results. Please do not forget to first remove results from the simulators folder before you submit them to avoid submission of duplicates. Duplicates will eventually be removed from the results database and from the high scores.
Rename results. To facilitate storing of results, you may rename the submitted results files according to whatever naming scheme you want. This has no influence on how good results can be processed. In fact, renaming might even make it easier to discuss specific problems that might arise.
Do not touch results. Trivial, but worth mentioning: You should not manipulate results. This includes opening results with text editors that might adjust line-endings or break up extremely long lines, before inviting you to save the mess to disk! Results are deliberately not kept secret, so you are welcome to look at them. But if you do, then please look at a copy of the untouched file and submit only the untouched file.
Submission intervals. You may submit results as often or infrequent as you want. Generally, it makes sense to roughly submit every few weeks. If you choose simulations that produce very large results-files, then you may also submit more often. As many email servers have problems to handle very large emails, you may want to make sure that your attachments stay below 2, 5 or 10 MB or wherever your respective limit might be. If you do not want to submit such large files, then simply choose tasks that take longer to compute. Apart from email limitations, file-size and the number of files submitted do not matter to evolution@home.

Finally, one last important point: Thank you for your participation in this adventure.

New project for Simulator005 looks at the combined effects of two types of mutations (1 Nov 2005)

Thanks to your contributions the first major project of Simulator005 has been completed to an acceptable degree. Now it is time to extend the very simple observations of mutation accumulation from the first project towards the more complicated case, where two types of mutations can potentially accumulate in the same population. This is an important step towards understanding more realistic situations as those faced by real organisms.

The first major simulation project of Simulator005 has been declared complete. With more than 83 years of contributed computing time and over 100 000 simulation results enough statistical power has been accumulated for the analyses that are being carried out at the moment. Those of you, who still want to contribute to that project can continue to do so and you should certainly submit the results you have already computed, as this will further increase the statistical power of future analyses. Some results from this project are being written up and others are under peer-review. As the scientific review of the results is beeing completed, you will read about it on this website.

Those of you who want to start new run-files can get them here. In these tasks Simulator005 looks at what happens, if two different Muller's ratchet processes operate in the same population at the same time. This collection of run-files contains 21680 simulations of an estimated CPU-time of less than 1 month with a total of 98.7 years most probable CPU-time and 714 simulations of an estimated CPU-time of more than 1 month with a total of 97.1 years most probable CPU-time. Since some of the run-files are quite long, some of you might want to consider editing them manually to compute some of the parameter combinations from the back sooner than their original order.

Muller's ratchet is already hard to understand if all mutations have the same effect; however, when two different types of mutations (occurring at different rates and having different effects) start to interact, then it becomes even harder to predict the rate of mutation accumulation. Since the estimate of the most likely computing time strongly depends on the rate of mutation accumulation, some of the predicted computing times will be very inaccurate. If that is the case and you counted on a simulation to complete at a certain date, then just stop the simulation. The incomplete result of that simulation will eventually end up in the results-file, so that it will be submitted with your next results submission. On the long run simulations like these will be used to improve computing time predictions.

The careful readers among you will have noted that this new project is the third, where as the last big project was the first. Between these two there was a small project that explored the upper limit of population sizes that can be computed on 32-bit CPUs. This project 2 contains only a limited parameter range of some very large simulations. Those of you, who enjoy exploring simulations with extremely large RAM requirements are invited to explore these run-files.

Software patents endanger global computing
(update 06 July 2005)

The general point of patents is to motivate innovation by protecting the work of inventors for a while. However, it turns out that software patents are really bad at fulfilling this purpose for many reasons. They can even be used to make comfortable global computing impossible.

You might think that software patents do not harm you - but that is only until the rightful owners choose to enforce them. Just imagine processing global computing work units without being allowed to easily check progress using a progress bar.
Yes, it is true that evolution@home could be improved significantly over its current state - without caring about software patents, and yes, we are working on the long overdue improvements. However in the mean time we would like to point out that software patents can easily complicate development or maintenance of evolution@home up to the point where it becomes impossible without the financial resources of a multi-billion-dollar company. Therefore, if you want to support evolution@home, you may well consider supporting initiatives that try to stop software patents. Visit www.ffii.org for more information. Sadly, neither news value nor magnitude of this problem have been decreasing over the past months. The recent decision of the European parlament to block the patent-friendly laws proposed by the European commission prevented a disaster, but the battle is far from over (see english and german). We still need laws that protect the freedom to write creative programs.

New forum available (25 May 2005)

Thanks to science@home, we now have a new forum in English and German.

The forum is open to all questions and discussion directly related to evolution@home, be it technical problems for new participants, highscores or future features. You can contribute to discussions there and perhaps you are even interested enough to help with the moderation.

Progress report (25 May 2005)

Quite a number of advances have been made under the surface of evolution@home. This small progress report brings you up to date.

Many of you have repeatedly requested such basic global computing features like full automation of task distribution, results submission and highscores updates. We have been working hard towards these goals and made considerable progress in building an infrastructure that will support this. The framework we developed includes

an infrastructure for transmission of input and output parameters between simulators and the server that schedules new simulations,
a database that stores all results as well as the tasks that still have to be computed and
a script that automatically generates redesigned highscores from the database without further manual interaction.

New dedicated hardware has been put in place and the database and server applications are being migrated to the new systems. This includes

Validation of results in the migrated database (quite a number of integrity checks have already been completed, but a few on the final production system are still needed)
Complete transition of all new submissions to the final database and switchover of highscore-production from the current limited system to the fully automated script that accesses the database. The first noticable change will be the new layout of the highscores and the fact that there will be one big united highscores list again, since the new system is large enough to keep all results in one place.
Merging of Simulator005 with the new fully automated parameter transmission system to eliminate the need to email results.
Transition of the static website you are currently reading to a future-proof content management system.

Progress has been slow, not only because of limited manpower, but also because of the complexity of the tasks at hand. Evolution@home is one of those global computing efforts that are more of the 'ant-hill'-type than of the 'ruby-in-the-rubbish' type. The latter just need to write results to archives and keep those few records at hand that exceed a specific score of interest. Ant-hill-type problems, in contrast, need to keep almost all results at hand, because they are typically used in combination to understand trends in the big picture. Thus they can be described as special data-warehouses that require considerable infrastructure to allow orderly access. Much of our energy goes into organizing the data at every level, from designing a new simulator (what will be interesting to include, what will just bloat databases?) to final results analysis (what should be stored, what is easy enough to be recomputed?). The goal of this process is, to eventually make simulation results publicly accessible for analysis by other scientists that do not have the computing power to address the corresponding models, but need their output to understand the organisms they are studying.

In addition to these tasks, development of the biological core of the next simulator takes up considerable energy together with the revisions necessary for publishing results of the current simulator. Actually, over the last few months most time went into development of an analytical model that will help check some results of the next simulator. The corresponding results are being written up for publication.

All core tasks that can be delegated under current circumstances have been delegated (DB+global computing, net-security and content-management) to voluntary coworkers. So please don't jump to conclusions, if things take a bit longer than in other projects that have a dedicated team of full-time staff.

Thanks to science@home, we now have a new forum in English and German. You can contribute to discussions there and perhaps you are even interested enough to help with the moderation.

Finally, we want to express a big thank you to all of you who contributed over 70 years of CPU time to what is already the worlds largest database for Muller's ratchet simulation results.

---

Please notice that this website will be migrated to a content management system that removes many of the current static limitations in the near future. Until migration is complete, update frequency will remain well below average.

Software patents endanger global computing
(update 05 May 2005)

Software patents endanger global computing
(23 Aug 2004)

A message from the initiator of evolution@home
(31 Mar 2004)

Many have wondered whether it would be possible to update highscores more frequently and why the website carries so little news. Some have even wondered whether evolution@home is still alive. To start with the latter, evolution@home is very much alive in the sense that participants are contributing results, the results are important and future improvements are under development.

To explain the lack of updates, one has to understand a bit more about the history of evolution@home. Basically, evolution@home started as a one-man-initiative and thus, all tasks that are easily distributed in other global computing systems (website development, writing news, generating run-files, generating highscores, developing code for the next client, analyzing results and writing scientific papers, etc.) had to be done by me with the inevitable result that each of these tasks took/takes correspondingly longer. In addition to that, my work requires/required quite a bit of biological research that contributes only indirectly to evolution@home (eg. estimate some interesting input parameters for future simulations like in L. Loewe et al. "High deleterious genomic mutation rate in the stationary phase of Escherichia coli" Science 302(2003)1558-1560). Currently I am

focusing the aim of the next simulator that will involve much of what Simulator005 is focused on, but includes recombination as well, an important extension, which has effects that are hard to predict. This requires solving a number of simple cases analytically to find the most interesting questions and then build Simulator006. That takes most of my time.
writing up biological results of the first round of analysis that was completed in 2002. Writing up means transforming the information into articles that are then submitted to scientific journals, which review them and publish them, if they can be improved according to the suggestions made by the reviewing scientists. This process is slow and takes the rest of my time. It is common sense among scientists that scientific results should be checked by the scientific community first before disseminating them in the general public. Therefore I will not post unchecked results-summaries on the website.

These two main tasks fill up my days and weekends and leave little time for updating high-scores and the website. Some of you have offered help, even for programming and I appreciate that very much. However, the evolution@home global computing system is quite complex and it is not easy to distribute development work, while the corresponding infrastructure is missing. Thus, at the moment the learning curve is very steep for meaningful work. Since I have to do all the explaining at the moment, I am very limited in my ability to train people. In the little time I have left for these issues, my main emphasis is currently, on building an infrastructure that makes it easy for other people to get involved. This requires

migrating the website to the typo3 content management system and an infrastructure that allows developers to discuss and document what they are doing, so that new developers can catch up by reading. The new infrastructure shall also facilitate production of local translations. I already delegated development of the typo3 infrastructure.
migrating management of single-run-results (and high-scores) to a proper database. Much of the necessary infrastructure has been developed together with a database specialist, but some additional programming and especially quite a bit of testing are necessary to go productive with that solution. Given that we are already thus far, it does not make sense to think about other ways of speeding up high-score publication, since the database and an automated client would lead to that goal eventually.

Thus, at the moment, I think that everything that I can possibly delegate is delegated already (the current semi-automated update system for high-scores is not easy to delegate; I would rather do it myself for the few times left in the life of the old system than explain the gory details to someone else). If you think that there is something that could be easily delegated to you, then you can send me an email explaining what it is that you want to contribute. Please include a description of the techniques you are familiar with and the experience you have along with the commitment in terms of time you are willing to give. This will help me to assess whether there is something suitable for you to help with now or in the future.

I know that evolution@home is currently missing a lot of exciting features found in other global computing systems. Unfortunately this cannot be changed over night, so it will take some time until evolution@home catches up. For those of you, who need such features now, I recommend any of the other existing global computing projects that offer the features needed. However, you might want to revisit evolution@home from time to time to watch it evolve.

Finally and most importantly, I would like to take this opportunity to thank each and every contributor. Simulations are crucial to the project and running them is one way that you can continue helping with evolution@home. Your involvement really does make a difference. With each simulation you are doing important work to increase our understanding of evolutionary forces, the details of which you will eventually be able to read about. I appreciate your continued faithful support over these years so much. It is fantastic to see how results drop in every day. Each one is like a small brush stroke that adds more detail to the modern picture of evolution. Thank you for your patience, understanding and continued participation.

Laurence Loewe.

Small but nice GUI-helper application released for Simulator005
(11 Apr 2003)

Michael Reich (Zahmekoses of Rechenkraft.de) has released a small nice graphical user interface for the semi-automated versions of Simulator005. It can be downloaded here (there is an english translation on the german site). This little program will work with all semi-automated versions of Simulator005 and shortens the waiting for the fully automated system that includes its own GUI.

This little program helps to control progress of the simulator and allows easier stopping for intermediate breaks. Thus you can stop computation of the simulator at any time (without closing the application!) to free your CPU from background activity. As soon as you want to continue background computing, you just have to bring the simulator to the front and select the command line menu item 1 (continue normally) to resume computation as if there were no break. While this is also possible without the GUI, it is now much easier and convenient to use. Furthermore, this GUI allows a quick visual overview over the progress of the simulation. Have fun.

Hottest run-files updated and highscores delayed (27 Mar 2003)

Please find the lastest set of run-files here. Publication of highscores is delayed due to technical problems.

After publishing the run-files from scheduling session 6, many results with low computational complexity were received. Thus, the corresponding runs were removed from the collection of scheduling session 6 to yield scheduling session 7.

The previous method of computing high-scores has run into trouble due to technical limitations. To cut a longer story short, some kind of transition has to be found in the near future and most probably it will be the transition to a high-score database.

New run-files are available (20 Mar 2003)

A new set of run-files has been produced with only minimal overlap to previously computed runs. These new tasks are immediately available for computation.

This new scheduling session focussed on situations, where the mutation rate is low and Muller's ratchet would therefore need more time to click - if it clicks at all. The current set of run-files (scheduling session 6) should be more uniform than the last one and should have better run-time predictions for most of the tasks.
The previous set of run-files (scheduling session 5) could be divided into tasks with quite precise run-time redictions and other tasks with completely off-the-mark predictions, the latter giving problems with high-score computation too. For those who wonder whether the corresponding high-scores have been lost, just a short note: No, your high-scores and the corresponding results have not been lost. They will be processed as soon as the corresponding problems will have been solved. However, this may still take a bit of time, as we are currently working hard to make a prototypic transition to full automation in the hopefully not too distant future (see the progress-page).

New publication about global computing with evolution@home background (7 Feb 2003)

A new publication about global computing and bioinformatics is available. The latest issue of Briefings in Bioinformatics contains a review paper that explores the possibilities of global computing for bioinformatics problems and sets this in context with grid technologies. It might be interesting for those, who want to get a broader perspective on the potential of global computing.

The paper under the title "Global computing for bioinformatics" can be found in volume 3 number 4 pages 377 - 388 of the Journal "Briefings in Bioinformatics" (Publisher: Henry Stewart Publications). Abstract: Global computing, the collaboration of idle PCs via the Internet in a SETI@home style, emerges as a new way of massive parallel multiprocessing with potentially enormous CPU power. Its relations to the broader, fast-moving field of Grid computing are discussed without attempting a review of the latter. This review (i) includes a short table of milestones in global computing history, (ii) lists opportunities global computing offers for bioinformatics, (iii) describes the structure of problems well suited for such an approach, (iv) analyses the anatomy of successful projects and (v) points to existing software frameworks. Finally, an evaluation of the various costs shows that global computing indeed has merit, if the problem to be solved is already coded appropriately and a suitable global computing framework can be found. Then, either significant amounts of computing power can be recruited from the general public, or - if employed in an enterprise-wide Intranet for security reasons - idle desktop PCs can substitute for an expensive dedicated cluster. Keywords: global computing; Internet distributed computing; Grid; in silico biology; simulations; structure prediction

Unfortunately, the copyright issues prevent distribution over this website unless a fee of 100 British Pounds is paid to the publisher. If this were not the case then this paper would be as available as the previous one. You may want to contribute to a discussion about fees in the publication of scientific papers.

Problem with high-scores (7 Feb 2003)

As some of the active participants have noticed, currently, a significant fraction of the results submitted is not being reported in the high-scores. This is due to a bug in the processing software, and not in the simulator. As all results are stored, the missing results will be included in the high-scores as soon as the bug is corrected.

In more detail: The problem occurs only in simulations, where Muller's ratchet clicks only extremely slowly. In this situation, the simple equation used to predict clicktime (and as a result of that CPU time) may yield numbers larger than what a 64 bit double precision float number can contain (eg. 10e400). The simulator correctly describes such numbers as "inf" in the result and thus the simulation is fully valid. However, the software that reads results and checks their integrity cannot deal with "inf" at the moment and therefore refuses to import such a result in the current database. As soon as handling of infinite numbers will be added to the results-checking code, all results that have not been imported yet will be processed. Thank you for your patience.

First scientific paper about evolution@home with IT-results is ready for download (24 October 2002)

While biological results are still being written up, the first scientific paper based on results submitted to evolution@home is ready for download now. It addresses some fundamental issues about global computing projects that span a large number of orders of magnitude in computational complexity.

For participants of evolution@home it has become natural to choose the complexity of their work units. However, this users-free-choice-system is new to global computing and it was not entirely clear from the beginning, how well it would work in practice. The first scientific paper about evolution@home adresses these and similar issues, like e.g. dependence of performance on CPU-cache structure. Another problem encountered was to quantify, how many orders of magnitude predicted and actual computing complexities differ on average (for releases 1-5 of Simulator005). As the assymetric properties of the commonly used relative error did not allow proper quantification, the "error of magnitude" was defined in this paper. Its values are symmetric around zero (= no error) and easy to understand intuitively (e.g. +1 is one order of magnitude, i.e. 10 times longer than predicted).

This paper was presented at the 2nd International Workshop on Global and Peer-to-Peer Computing on Large Scale Distributed Systems at the IEEE International Symposium on Cluster Computing and the Grid (CCGrid'2002) in the Berlin-Brandenburg Academy of Sciences and Humanities, May 21 - 24, 2002 Berlin, Germany (see http://www.ccgrid.org/ and http://ccgrid2002.zib.de/) and has been published in the corresponding proceedings: Loewe (2002) "evolution@home: Experiences with work units that span more than 7 orders of magnitude in computational complexity", pp. 425-431 in: Bal et al. (eds) Proceedings of the 2nd IEEE/ACM International Symposium on Cluster Computing and the Grid (CCGrid2002), Berlin, Germany, IEEE Computer Society. You can get it free under URL http://www.evolutionary-research.net/Science/Papers/2002/Loewe2002-EaHworkunits.pdf.

As no high-scores existed at the time of the writing of the paper, the following acknowledgements are added here: RabeRudi, Paranoia Retnek, arswitchman switchman and many others made significant computing contributions as participants of evolution@home. Without their help it would not have been possible to conduct this work.

Please update to release 6 of Simulator005 and get new tasks from the run file scheduling session 5 of project 1 (07 October 2002)

Intense analysis of submitted results led to significant insight that allowed better predictions of computing time for Simulator005 and to detection of a bug that affected a fraction of the results computed up to now. The new release 6 implements the corresponding changes along with other minor improvements. Simultaneously, a new collection of run-files with completely new tasks was published. Please update as soon as possible.

After a long waiting period a new release of Simulator005 has finally arrived. The improvement of predictions of computing time applies in particular to situations, where Muller's ratchet hardly ever (or never) clicks because all mutations are removed by purifying selection. The earlier simulators could easily miss such tasks by several orders of magnitude. The bug mentioned above affected the generation of Poisson random number derivates and introduced a systematic downward bias, but only for simulations with a mutation rate of less than 0.001 (the smaller part of all simulations that were run up to now). To eliminate the last chance that you continue with such a task, please stop all simulations now and update the simulator.
A new system to generate run-files was implemented as well. It comes not only with a handy table that helps you to choose the RAM complexity of a simulation, but now you can also easily choose the computing time complexity. While tasks are no longer sorted according to CPU time complexity within a run-file, they are now arranged according to fixed categories that help you even better to choose files that suit your computing habits. As the system is designed to easily add new scheduling sessions, you may see more frequent changes in the collection of hottest run-files.

High scores for evolution@home improved again
(17 August 2002)

Now the high scores of evolution@home include total summary statistics.

Total summary statistics are one of the interesting points in every global computing adventure. Often, only the total number of work units and their overall CPU-time is given. In addition to that, the high scores of evolution@home list the total amount of work (in GigaIndividuals), the average duration, performance and memory requirements of simulations, the performance that a supercomputer would need to compute all results submitted in one day and the amount of RAM, if it would need to compute all completed simulations simultaneously. This information will help to set individual contributions into context and facilitate comparisons with other global computing projects.

The first high scores for evolution@home are available
(6 July 2002)

Finally, evolution@home has started to publish detailled computing statistics, that list the contributions of non-anonymous participants.

As the work-units available for computation at evolution@home have a extraordinary diversity, several high score lists are available (e.g. total work, total CPU-time, longest single run and maximal performance among others). Thus, there will allways be more than one top-contributor, since you can not run the longest simulations and yet contribute the highest number of single runs. Therefore, pick your field and stay there, if you want to reach to the top.
Updating the current high scores is semi-automated. Therefore, updates will be made about every two weeks.

The unofficial forum of evol ution@home is up and running
(1 July 2002)

Now evolution@home has its own discussion forum. Everybody is invited to discuss topics related to the ongoing efforts of the first global computing system for evolutionary biology. From computing milestones of your team, over things you like or do not like about evolution@home, up to future directions of the system, your questions, contributions and opinions are welcome. Thanks to David Robinson, now you can share your thoughts and computing experiences.

Welcome to the unofficial forum of evolution@home! This is a great place to be, when you are interested in evolution@home and want to be part of the community of those that dedicate their computers to do evolutionary research.
In the forum you may read about some of the latest news, discuss anything regarding evolution@home, like e.g. simple how-to questions, technical aspects on operating simulators and contributing computing time, technical features that you would like for future simulators, general things you would like to change, things you would not like to change, future directions of evolution@home, ...... You may also introduce your computing group, discuss progress in high-score races and report computing milestones of your group.
To keep its focus and maintain reasonable quality, the forum is moderated. You are expected not to personally attack people or promote lies, high-score cheating tricks, other hacks or illegal things, post excessively repeated solicitations of becoming member in your team or discuss anything that has nothing to do with evolution@home, be it of general biological, evolutionary, philosophical, world-view related, religious, political or other interest.
Now, have much fun in The unofficial forum of evolution@home!

evolution@home continues and needs many more stochastic repeats than anticipated (1 July 2002)

Although the evolution@home website had not been updated for a long time, evolution@home is up and running and will do so for the coming years. The technical issues that prevented a more frequent update have been solved now. Preliminary analyses have shown that a higher number of repeats for each simulation is necessary to determine the shape of distributions of interesting output parameters with reasonable accuracy.

As the website of evolution@home had not been updated since more than 9 months, not few people were wondering, wheter evolution@home is still active. To make it plain: evolution@home is a long-term global computing system that targets a long series of computing projects for a long series of simulators under development.

However, as manpower at evolution@home is quite limited up to now, things take their time. The largest problem that prevented more frequent updates was a long needed transition from one web-design software package to another, combined with very urgent other research priorities.

Before other projects can be started or major global computing feature development will go on, the more than 23000 results collected up to now (> 12 years computing time) have to be analyzed. Considerable progress is being made in this. One result has immediate consequences for further computations and, therefore, shall be mentioned here: The distributions of clicktimes of Muller's ratchet as measured by Simulator005 have considerable variance. To further quantify these distributions in detail, many more stochastic repeats of simulations are needed. Thus, the run-files that can be currently downloaded from the web will remain interesting for quite some time and their repeated computation is highly welcome, as each new start of such a simulation automatically produces a new result by generating a new random-seed.

Successful prototypic analysis of first evolution@home results encourages further computations (14 September 2001)

The first glance at the results of evolution@home is promising. Despite its tedious, semi-automated way of operation, more than 120 individuals contributed nearly 5 years of CPU-time from April to mid August 2001. Some of the run-files have been removed from the website, as they are completed now.
Until an automated priority setting system for run-file production is in place, please help compute simulations with longer computation times (less work for you and your modem, more work for your computer).

At 13th of August the cut for the first (still manual) analysis session was made. All results submitted by email until then were included in the analysis made for a poster at the Eighth Congress of the European Society for Evolutionary Biology (Laurence Loewe, "Predicting extinctions due to Muller's ratchet in humans and bacteria", 20-25 August 2001, Aarhus, Denmark). Details on the biological meaning of the results will be published in the coming months. While it will still take some time until analysis of results submitted will become fully automated (including high-scores computation), a first assessment allowed to estimate some global statistics. More than 500 results mails with more than 100 MB of results were recieved by more than 120 participants.

The combined, sustained number crunching capacity of Simulator005 in that time was about 2 million individuals per second. This equals to nearly 10 fully dedicated up-to-date PCs running continously. While this seems little compared to other projects that are fully automated and have been well advertized, this is remarkable given the amount of manual interaction involved, the virtual absense of advertisement and the lack of high-scores that ensure every participant the project is still active. To answer a question many may have thought and a few have asked: evolution@home is active and will remain active even if it takes some time until the website is updated . There are no plans to stop it and if this changes, it will be announced at this website. All results submitted will be evaluated manually from time to time and automatically once the corresponding software has been developed. After that updates of the website will become much more frequent.

New release of evolution@home simulator no longer interferes significantly with business as usual on a PC (19 May 2001)

Where earlier releases slowed down other applications considerably as they were running with normal priority under Windows, the new release 5 of simulator S005 does this no longer. Thus you can use your PC for graphics intense applications as ever without even noticing the evolutionary research done in the background.

To check the CPU usage of the simulator and of other processes under Windows, you can use Microsofts WinTop utility or the task manager on Windows NT systems. Doing this may also help you uncover unwanted activities on your system.
While the structure of the MacOS classic does not allow such simple priority setting due to its cooperative multitasking model, the new release is made to be as cooperative as ever. With the new default values you will hardly notice the simulators operations during normal work. However, if you do, the MacOS simulators are easier to stop non-destructively (just press a key) than the Windows counterparts. As CPU usage on a Mac can be set in the preferences, you have the full control over the CPU-time you dedicate to evolutionary research.
Further improvements include a new name for the results-file making it easier to attach it to an email using certain email-programms.

First manual evolution@home simulator becomes widely available for Windows and MacOS to help uncover genetic causes for extinction of species. (27 April 2001)

You have been caring about endangered species? You are generally interested in evolution of life on earth? You are a specialist for the evolution of sex? Then you will want to download the new simulator of evolution@home. It allows your computer to do meaningful evolutionary research while you have a cup of coffee or enjoy the weekend.

This first simulator targets a phenomenon known from population genetics: Muller's ratchet in asexual populations. The essence of it is the stochastic accumulation of slightly deleterious mutations in genomes over evolutionary time. If this happens in a species, it might be driven to extinction not due to environmental, but due to genetic reasons: the best genomes available become increasingly contaminated by mutations that are harmless, if rare, but dangerous, if frequent in a genome. Of course Muller's ratchet does not click everywhere. Evolutionary geneticists have been trying to predict the rate (and thus long-term effects) of the Ratchet for more than 25 years now. They found this problem to be very resistant against comprehensive analytic solutions. In the mean time some predictions can be made with good accuracy. However, large ranges of parameterspace are still hard to tackle.

This is where the current simulator comes in: by systematic scanning of parameterspace our understanding of Muller's ratchet will grow. Ultimately the new simulator will produce a database with simulation results that represent thousands of years CPU-time. It will help biologists to easily estimate the effects of Muller's ratchet for the organisms they work with every day - without having to understand all the technical details of the predictions. This database will also help theoretical population geneticists to develop better predictions of Muller's ratchet by comparing their theories with a large collection of simulation results. Finally, comparing the data from these parameter space searches with the parameters found in biology, will help estimate the extent to which genomic decay by Muller's ratchet indeed does contribute to the extinction of species. To understand it is essential for fighting it.

The current simulator that allows you to participate in this adventure is now available in a manual version for various flavours of MacOS and Windows. As starting the project had a higher priority than implementing full automatization, the simulator-software comes as a console application that only writes and reads to the folder it is located in. This means that you have to provide the input (in form of a run-file) and you have to submit the results (in form of a results-file emailed as an attachment to simulator005@evolution-at-home.net). However, the amount of manual work is minimal, especially if you choose run-files that run for weeks or longer. Detailed instructions are included with the simulator. Besides having your computer run when you would normally switch it off, there are no hidden costs if you want to participate (no automatic data-intensive dial-in connections to the Internet).

But what do you gain? We can not offer you to pay for your contribution or give you the chance to win 1000$, as this is a non-profit research project. (If you want money, go to one of the other global distributed computing projects that do it, but be warned: not many people have become rich that way.) However, we do invite you to personalize the results you compute. Thus you will get an entry in the high scores lists that will be published (including the possibility to set a link to your home page). However, most interesting will be the fact that you help mankind to understand a part of the causes that threaten so many species on our planet. Thereby you become part of those who do something to save lifes.

So, make up your mind and join us in the adventure of searching for the mysteries of the evolution of life on planet earth.

Evolution@home has started:
The first simulator is available for MacOS (3 April 2001)

As eagerly awaited, evolution@home has finally started. You can download the first simulator targeting Muller's ratchet in asexual populations. Then select those simulations that fit your computing habits from a wide range of parameter combinations (=keys) with different RAM and CPU-time requirements. Finally, please submit results by email. We hope you enjoy evolutionary-research and have a pleasant computing experience.

This simulator helps uncover potential genetic causes of extinction for endangered and not-yet-endangered species by investigating Muller's ratchet. Your help to improve understanding of such genomic decay might one day be used to fight it. Watch out for more detailed information, as well as a Windows-version of the simulator in the next days.

Change in strategy will make first Simulators available soon
(1 March 2001)

The development of the next version of the framework Eepsilon will need longer than anticipated. Thus, to solve the questions at hand, existing code is being revised to allow distributed simulation. Therefore evolution@home will start in the next few weeks for those who donât mind getting some keys from a website and email results back to evolutionary-research.

The core strategy of evolutionary-research remains to develop Eepsilon as a framework for solving problems in evolutionary biology by simulating populations. Global computing naturally comes in, as the complexity of these tasks is enormous. Much energy is being spent on making the framework as flexible as possible in order to lay a solid foundation for long term usability. This did and will need time.
In the meanwhile, however, some have been asking when evolution@home will finally start and at evolutionary-research curiosity has been increasing: While some limited simulations on Muller's ratchet have been conducted by now, we simply lack the computing power to solve some very interesting questions.
Therefore a simple decision has been made: Development of the general framework will have to wait for the moment and the code already available from the last design version of Eepsilon will be used to tackle the questions at hand. This means, however, that the benefits of the new framework will not be available for these first simulatiors.

Thus, practically, in the next few weeks you will be able to

download the programm and put it into a dedicated folder on your hard disk.

look at the website of evolutionary-research to get the keys for the computing tasks you are willing to contribute to (choose from many options regarding computing time, RAM needed and size of results-email.)

put the keys you selected to a text-file (as simple as copy and paste from the website)

save that text file under the name "run"and move it to the dedicated folder on your hard disk

start the programm (and enter your name, if you want to personalize the results you compute). The programm will automatically read the ãrunä file and write the results to a file called "results.txt"in the same folder. (It will also produce a log file with all screen output)

allow each key to run until it terminates, as saving of intermediate states is not possible in the current version. So pick the right computing time when selecting the keys.

send results back to evolutionary-research by sending an email with your ãresults.txtä attatched to the corresponding email address you will find on the downloadpage.

The amount of manual work involved is not as bad as it might sound, since you can handle many complex computations by only one cycle of organizing input and emailing results back: You will just put all the keys in one run-file. However, people who want it completely automated will have to wait for the new version of the framework.

German website of evolutionary-research online
(4 Dezember 2000)

The geman translation of the evolutionary-research website has gone online. We wish all our readers from Germany a pleasant start with the second global distributed computing project initiated in Germany (Xpulsar@home from TŸbingen being the first).

Development progress tracking becomes public
(1 November 2000)

Development at evolutionary-research has been going on all the time since finalizing the cornerstones of the Eepsilon design in June 2000 and putting the website up a month later. Some have been wondering about the current status of the project while nothing new appeared on the website. Now you can find a rough status report on the developers progress page. So while you will have to wait for some months before downloading the first client, you can at least track progress of development, if you want.

evolutionary-research announces distributed computing project evolution@home (1 August 2000)

evolutionary-research has started to inform the public about an upcoming grand - challenge computation research program aiming at the simulation of evolution. A major focus of the project called evolution@home will be the simulation of effects of various evolutionary factors

on the survival of populations of endangered species and

on the evolution of novel functional adaptations.

This will be the first possibility for the general public to participate in serious scientific evolutionary research. The project will follow the successful pattern of similar global distributed computing enterprises like the SETI@home project: You will be able to download a client-software that runs on your computer and calculates the evolution of models that had been scheduled for that on a central server. By doing this you contribute to a better understanding of evolution. The results obtained will be published. See the introduction. If you want to participate, you can sign up now, to show your interest and recieve email when the free client software become available.

evolutionary-research is now online (8 July 2000)

evolutionary-research has launched its new web-site. Over the coming years it shall develop into a valuable resource for all those who are interested in serious evolutionary research. This site will report progress made in the grand search for the effects of evolutionary forces and will also include some more general interesting news related to evolutionary biology. Helpful links, selected nice pictures and scientific background information will be included too. So bookmark this page and return from time to time to watch its evolution.


	© by evolutionary-research, last change 2006-09-14 . Contact


	[Home] [Info] [Simulators] [Scores] [Science] [Results] [Discuss] [Archive] [Search]