Message from Ant.
I mentioned many months ago - some time in early 2015 - that GridRepublic in Boston offered to take over the server management of FiND@H.
It was - and still is - my intention to make results from the FiND@H project publically available. I've built a website that displays the first results - http://chemware.ucd.ie/FiNDaH/.
The idea is that anyone, anywhere, has access to this data and can validate the results in the lab.
That's where things get tricky. If an academic lab - or for that matter a biotech company - puts their own time and money into validating these data, then they get the patents and profit. Ideally government money should pay for these validation studies - socialising the profits - but as I am in Ireland and computational drug discovery is off the bottom of the priority list, I repeatedly failed to get funding for the validation studies from the HRB, SFI and EI research funding agencies.
I am now no longer employed to be a research scientist and have been forced to take an admin role at a college to break out of the annual contract life of postdocs. Hence the project went dormant until GR picked it up again.
But I still want the results validated in the lab. I repeatedly tried to get funding from the Gates foundation, but ironically they refused to fund projects that where computational. I tried croudfunding, and the generous donations we received covered the cost of new hard drives - a far cry from the 100,000s needed to support lab validation studies.
So I'm still hopeful that some angel investor will swoop in and offer to validate the work. Will the profits be shared? - no. My contract with the FiND@H volunteers ends with porting the data onto the web. Now anyone in the world with deep pockets can take advantage of the data. The risk is still high and there is no final guarantee. Of course I hope I'm right, and that this project works. But the person who puts hard cash down to validate the data is the one who walks away with the spoils. At the moment this could be a lab/company in India or South Africa or wherever... for all I know.
I'm just delighted that GridRepublic have offered to keep the project alive... I think we could use this method for all sorts of infectious diseases, given the chance. In the end I hope someone, somewhere, puts in the effort to validate the data in vitro.
Ps: I have not heard of any BOINC project socialising the profits from the computation work. For instance, the WCG project similar to ours - Go Fight Malaria At Home - also put all their data online. I have no idea if this has been validated in house by them - then they could hold patents on good results and license to large pharma and (potentially) make profits from it. I have no idea... just saying. The nett result of this work are possible hypothetical interactions, which will take a lot of time and money to validate.
Also, look here -> https://boinc.berkeley.edu/dev/forum_thread.php?id=11309 7 Dec 2016, 9:02:52 UTC · Comment
If the website stop working suddenly it's because I'm moving the database to a new server.
But I've no idea when.
:) 13 Jan 2016, 13:21:46 UTC · Comment
The BOINC database, the one which stores the WUs, the Users, the forum, etc. crashed. But I wasn't sysamdin on this server.
We rebooted the server and I'm now going to have rights on it. 27 Oct 2015, 14:38:23 UTC · Comment
I don't know why but the server just crashed this weekend. Not BOINC but the entire system.
I had to cut off the power and reboot everything. 14 Sep 2015, 14:03:27 UTC · Comment
Due to an XSS issues in server_status.php, I updated it.
PyAssimilator.py (vina) works well, do not mind the red warning.
I'm a little bit stupid, I forgot to backup my code for the progress bar... I'll do it again but maybe not today. 26 Jun 2015, 12:22:42 UTC · Comment
I have to shut down the server for the weekend. UCD is going to do some tests with the power in the building where the database is hosted. 5 Jun 2015, 11:56:57 UTC · Comment
Damn that's big :)
It "should" work. For now I just created 10,000 WUs, to test the estimation. 12 May 2015, 10:08:38 UTC · Comment
I said it will be 60M WUs but we have some work to do before.
Happy crunching! 20 Mar 2015, 11:50:42 UTC · Comment
Back to normal
The results database crashed because the server was out of memory...
Some experiments, 18pstiff2re for instance, contained many bad ligand structure. That's why you all had many failed WU.
Plan for the future.
After this batch, I'll wait about one or two weeks in order to get the last remaining WUs. Then, I'll extract all the results, clean the database, clean the server and insert 60 million WU into it. It's a very huge batch so I want to start from a empty database. 3 Mar 2015, 10:18:57 UTC · Comment
News about me
I just want you all to know that this project was a 6 months project that was part of my PhD programme.
Now, I have to work in another lab but I'll still maintain this server, because I really like what we are doing here. But I'll may be less present.
I would like to clean the database, and the server before inserting the HUGE batch of 60M WUs. It'll keep you busy for the next couple of weeks :)
I'm now working with Rhodococcus equi, that causes a TB like illness among the foal. It's a wet lab based project. 26 Feb 2015, 11:37:40 UTC · Comment
After the huge crash of the database, everything is fixed and ready to go.
2,906,889 more WUs!
Edit: The transitioner need to catch up 23 Feb 2015, 16:42:17 UTC · Comment
Project down this weekend
I stopped the BOINC server. I'm currently putting 3 million task into the database. The script will take some time. I'm going home now. I will restart the server on Monday. 20 Feb 2015, 18:17:22 UTC · Comment
I increased the predicted running time for this batch. Some WU may take some time to complete. 3 Feb 2015, 11:41:44 UTC · Comment
I just put 3 192 552 more WU!
Have fun :) 13 Jan 2015, 11:23:52 UTC · Comment
FightMalaria@Home data now available online
Hi all volunteers,
We have made a results website for easy access to the data we generated using the FM@H BOINC server. The first 400 MMV compounds (MalariaBox) data is available here:
We're still doing the last repeat calculations for the remaining MMV19k and other datasets, the results of which we'll add to this website. We'll also add the TB data to this site, once we've cleaned it up and added orthogonal scoring data.
I had hoped to get the publication out prior to releasing this 'raw' data, but this is taking longer than I'd hoped. Once the paper is submitted I'll add a pre-print to the website also. The scoring functions will then make more sense.
Thanks again for your generous contributions to this project.
We're optimistic that someone, somewhere, will use this data to confirm our findings and provide the world with novel drugs... and new targets for further anti-malarial drug discovery.
Ant 6 Jan 2015, 16:35:07 UTC · Comment
Waiting for data. And checking TB ones.
The number of WU computed since the rebirth of the project is 6 382 787.
I will need some time to check everything, and maybe create new WU for weird results.
There are only 22 000 errors. That's 0.3% :)
Well done. We need to create a new database in order to work on a new disease.
Wait and see but be ready!
Android bin (if possible)
(please comment to add stuff here) 6 Jan 2015, 15:46:20 UTC · Comment
FightNeglectedDiseases@Home server is up... and almost running
The long awaited revamp of the FM@H server has finally happened - thanks to Benjamin. And thanks to the generous donations that we collected at the http://www.gofundme.com/1l0u2o website we were able to buy another SSD to build the FiND@H server on. This should be super-fast and will allow us to venture into different diseases beyond malaria, using the same 'whole genome docking' approach.
We should have the first work-units ready within a week or two.
Thanks again for your patience.
Ant 26 Nov 2014, 18:40:30 UTC · Comment
We're still chipping away at the data...
Sorry for the delay in responding to your messages.
We have compiled all the data into a single database (no mean feat as it's now 118GB!) due to our limited resources this was spread over a number of HDs. The FM@H data is 95% complete. Kevin is in the process of pulling out the remaining failed dockings for re-docking on the BOINC server.
The new name of the server will be FiND@Home for FightNeglectedDiseases@Home. Over the summer we'll build a website front-end for the data, that will be searchable and interlink with other online resources (e.g. Uniprot/PDB/KEGG). This should make the data easily accessible for biologists interested in exploring the output data. We'll build it with future BOINC data from other genomes in mind (e.g. TB).
Many thanks for hanging in there for us...
ps: We would appreciate it if users do not bad-mouth other projects, as we do not want to have to act as social police. The reason we're NOT concentrating on these boards at the moment is because we're trying to concentrate on getting the data we have into a presentable format for publication. And to focus on validation studies.
14 Mar 2014, 10:56:43 UTC · Comment
Errors in current batch or WUs - BACK ONLINE
I'm sure you have noticed errors in the latest batch of WUs. This is due to formatting errors on a few ligands and results in many failed WUs because each ligand is paired with each receptor, therefore 1 failed ligand is muliplied by the number of proteins with which it is docked.
We have always had some errors in our ligands but the rate has been quite low (< 0.5%) overall until now. As the rate has increased (8% in the last 24 hours), I will be disabling the work generator until tomorrow afternoon while I weed out some of the errors.
Thanks for the feedback,
Kevin 18 Nov 2013, 18:46:47 UTC · Comment
We had a power outage over the weekend. The server is back up now and we should be up and running with new WUs in the next 30 mins.
Kevin 11 Nov 2013, 11:02:54 UTC · Comment
No new tasks for 24 hours
Sorry guys but the server is overloaded at the moment. We have stopped submitting new tasks while the daemons catch up. Also, we are going to start a new experiment tomorrow afternoon. Hopefully we will be up and running by 5.
Kevin 4 Nov 2013, 17:57:50 UTC · Comment
Brief interruption to supply of tasks
We are preparing the next experiment now. We expect to have more tasks this evening.
Kev 1 Nov 2013, 10:16:01 UTC · Comment
Short Downtime - UPDATE - Back online now
I will be upgrading some hard drives today so the server will be out of action for a few hours. This is the beginning of our upgrade work though it will be a few weeks before we have the new servers up and running.
I'll keep you posted.
Kevin 22 Oct 2013, 9:34:37 UTC · Comment
Another new server arrived today!!!!
We're just amazed at the generosity of the BOINC community out there. And no better evidence than the new server sitting on my desk today. A huge thank-you to Raymond Ottusch and Jeremiah Campbell who donated a server to the FightMalaria@Home project. The added complication is that it needed to be shipped from New York to Dublin at the cost of €350, which Ray generously covered!
And another thank you goes to Bobby Mitchell from Pure It Recycling (www.pureitrefurbished.co.uk) and Michael Mills from Spinnaker Waste Management (www.spinnakerwaste.co.uk) who donated the other server, organized by Tom Leggat (www.qubii.com).
And we've used some of our GoFundMe donation money to buy an SSD drive to insert into our existing results server.
So within a few weeks you should notice a massive change in our capacity to deliver and process workunits.
Again, thank you, thank you, thank you!
11 Oct 2013, 13:33:55 UTC · Comment
Thus far we have completed 114,480,481 results of 193,373,736 planned docking experiments. 59.2% completion.
Exp_all Exp_compared Results % completion
Xray_lig 24,288 24,288 20,944 86.2%
MMV400 1,590,700 1,590,700 1,556,650 97.9%
MMV19k 123,580,044 123,580,044 86,772,129 70.2%
NPC 9,231,196 9,231,196 9,008,621 97.6%
ChEMBL 47,662,569 47,662,569 17,122,137 35.9%
liver 11,284,938 11,284,939 0 0.0%
Total 193,373,735 193,373,736 114,480,481 59.2%
So at the moment we still have 78,893,255 docking calculations planned. Thereafter we'll add TB targets into the pipeline.
Thanks to all for your continued support with this project.
24 Sep 2013, 16:47:38 UTC · Comment
New server arrived today!
Many thanks to Tom Leggat at www.qubii.com for organizing the donation and delivery of a reclaimed blade server (Dell PowerEdge2950) that we'll install as the new front-end for the FM@H project. And thanks also to all those who donated to our fundraising appeal, as we'll now use that money to buy hard drives for this new server.
Thanks you, thank you, thank you!
17 Sep 2013, 15:15:33 UTC · Comment
Downtime - until Monday evening (~19:00 GMT+1)
We are currently between experiments and we are working on the next experiment at the moment.
We will have new tasks ready to send on Monday evening (~19:00 GMT+1). We apologize for the latest string of issues and will ensure we don't have a repeat of them.
Thanks for contributing and we hope you stick around,
1 Aug 2013, 18:53:41 UTC · Comment
Several people have kindly offered to donate hardware to this project. We very much appreciate this.
Unfortunately there are two issues with this:
1. The hardware must be located in our server room for security reasons.
2. Our IT department requires that all servers in the server room are less than 5 years old. It's their way of reducing the number of hardware failures and using the space in the server room as best as they can.
What we're looking for:
Front end server:
This server runs the BOINC daemons and hosts all downloads/uploads. Ideally this would be a reasonably fast multicore server with a fair bit of RAM (>16GB).
BOINC database server:
We already have a good server for this but it was loaned to us and may be recalled at any time.
HP Proliant DL385 G5,
2 Quad Core processors (AMD Opteron 2356 @2.3GHz),
32 GB of memory
If you are interested in donating, please do contact us.
Kevin 1 Jul 2013, 11:10:08 UTC · Comment
Changes to credit system
As you may have noticed, we have had some problems issuing credit with the latest experiment and runtimes have increased massively. Credit is currently granted based on a multiple regression which takes into account several factors that determine runtime. I have slightly increased the amount of credit granted per workunit as people have suggested that it it too low. Also, we have removed the very long tasks which should result in fairer issuance of credit. Please note that there will be outliers but on average the amount of credit awared should be proportional to the work done.
Kevin 1 Jul 2013, 10:36:30 UTC · Comment
GoFundMe server appeal
After a recent suggestion from BOINC users (http://boinc.ucd.ie/fmah/forum_thread.php?id=93), we're going to attempt crowdfunding for a new server to help us cope with the increased load now that ~16,000 computers are connected to our project in 127 countries!
If you'd like to contribute, please visit:
Ant 11 Jun 2013, 11:05:59 UTC · Comment
No new tasks for 24 hours
The server is overloaded at the moment. I'm going to turn off the work generator for 24-48 hours to let the assimilator catch up.
Hopefully it will be back up before the weekend is over.
Kevin 24 May 2013, 15:57:03 UTC · Comment
Exciting news from the FM@H project - first in vitro validation comes up with hits!
We've had our first major success from the FightMalaria@Home project. After analyzing the data generated in the first phase (docking all the 400 MalariaBox compounds against all models), we have ranked lists of potential inhibitors for each receptor. But are these real? This is the kind of question that keeps us up at night.
Now we can rest easy - we've tested the top 20 hits using a model of the Plasmodial form of HDAC (histone deacetylase) in a laboratory (Dr Marian Brennan, RCSI, Dublin), and found THREE new inhibitors! So that's just one of the 1,500 receptors that we're docking, using only the best docking results (i.e. only the 20 best ligands).
It's a fantastic result, giving us hope that there are other gems waiting to be found in the data that you've generated on your home PCs!
Well done everyone!!!!!
ps: We're in the process of writing up a manuscript for publication in an open-access journal, so you'll be able to read about all the details, and download all the data generated from FM@H, once we've survived the peer review process.
16 May 2013, 9:41:49 UTC · Comment
No new tasks for 24 hours
We are going to turn off the work generator for 24 hours to clear the backlog of results.
Thanks to everyone who participated in the challenge.
Kevin 2 May 2013, 17:34:49 UTC · Comment
One of the tables in our task database has been corrupted. As a result we will be down for at least a few hours while we restore from a recent backup and ensure that we aren't missing anything. We will be back up and running soon.
Kevin 10 Apr 2013, 12:50:31 UTC · Comment
We found a small bug in the assimilator daemon which prevented assimilation of WUs for about 2.5 days. We are now turning off the work generator to allow the assimilator to catch up. We will likely be up and running again at this time tomorrow.
Kevin 2 Apr 2013, 16:46:33 UTC · Comment
New workunits in 24 hours
We will be deploying new workunits in the next 24 hours. Between now and then you will not be able to receive new WUs.
Kevin 24 Mar 2013, 18:05:44 UTC · Comment
A few broken WUs
We are cycling through some old WUs some of which contain errors. There only enough for about 12 hours of computation and things will be back to normal thereafter. This is just to finalize a set of completed experiments.
Kevin 9 Jan 2013, 17:25:57 UTC · Comment
There will be a short downtime today due to a hardware fault. All currently running tasks will be completed but new tasks will not be issued. I hope to have everything running again in the next few hours.
Thanks for you patience,
Unfortunately it's going to take a little longer than expected. We will be back up and running in 24 hours.
Kevin 12 Nov 2012, 12:41:49 UTC · Comment
Changes to the credit system
We are currently working on a new credit system which will ensure that everyone is awarded a fair amount of credit. The current system is failing as some users are receiving far more credits than others for the same amount of work.
To fix this we will predefine the amount of credit for each workunit. The amount will be based on a multiple regression of previous results from my own trusted hosts. It will likely be ready early next week.
Thanks to everyone who has pointed out these problems.
I have hidden all other posts in this thread. Most posters have not broken any rules so please do not take offence this. We are switching to 1 credit per WU later today while we wait for the new system. 23 Oct 2012, 10:50:09 UTC · Comment
No new jobs for 24 hours
We need to clear out the backlog that has accumulated in the validator, so we've stopped new jobs for a while. As soon as the server has caught up, we'll set new jobs going again. And early next week we'll purge the DB, so the server status should be back up and running again.
Thanks again for all your help and comments.
5 Oct 2012, 11:03:14 UTC · Comment
Interview with Ant & Kevin
For those of you who want to see whose behind the icons:
http://www.ucd.ie/casl/newsmedia/caslmedia/title,135839,en.html 3 Oct 2012, 14:37:41 UTC · Comment
2 Oct 2012, 12:54:42 UTC
We have collected the experiments into groups, each aimed at a separate paper to be submitted to an open-access journal in the near future.
We have over 4 million results thus far, and % completion is as follows:
Progress has slowed as we're now docking larger molecules with more rotatable bonds.
2 Oct 2012, 12:54:42 UTC
We're back online now. ~1.5 million dockings ready to go.
We still have some improvements to make and we will be making these over the coming weeks.
Thanks for your help!
Kevin 24 Aug 2012, 15:40:41 UTC · Comment
We're still having problems with the server. However, the good news is research IT have agreed to provide us with two far better servers. We hope to move the project to these over the weekend but it will likely be Wednesday before we are back up and running. We will not be sending new workunits until then.
Thanks for you patience,
Kevin 17 Aug 2012, 17:19:27 UTC · Comment
We have had some server problems. For this reason, we will not be sending new tasks for a few days.
Sorry for the inconvenience. We will be up and running again soon. New apps and workunits are coming soon too. The new apps should be a bit faster and compatable with a wider range of hosts, while the workunits will be significantly shorter.
Kevin 13 Aug 2012, 15:46:54 UTC · Comment
Test experiment A is finished!
We have the first set of 144000 data points to sift through from our 'control' experiment. The system seems to be working well. The next batch of 1.2 million docking calculations is in the queue and ready to go. 27 Jul 2012, 23:33:07 UTC · Comment
FMaH server will be down for about 12 hours
Sorry for the inconvenience, but we will not be sending out new tasks for the next 12 hours. We're still optimising the system and will need to go off-line occasionally until Monday 30th July. 25 Jul 2012, 15:23:52 UTC · Comment
We have a few updates:
1. We have regenerated our signing keys but due to a technical problem they will not be distributed automatically. Please detach and reattach to the project in order to receive new workunits. We apologise for the inconvenience and will ensure that it does not happen again.
2. We are having issues with a small percentage (less than 1%) of Windows computers and a very small percentage of Linux computers. We are working to release new app versions in order to address these problems.
3. We have limited the number of tasks in progress per computer to 20 per CPU core. Although this has not been an issue for some, we have had a large number of aborted tasks as a result of sending too much work to each node.
Thanks for your your help.
Kevin 23 Jul 2012, 17:45:22 UTC · Comment
FM@H up and running
We have just fired off the first 144,000 docking runs.
If you'd like to see the details behind the 1,000,000,000 docking runs planned, please go to the 'Idea' page on our website:
After the first 94 million docking runs, we'll have achieved our first major milestone.
We may have a small delay next week between experiments, then it's all systems go.
Thanks again for your your help with this.
20 Jul 2012, 18:08:58 UTC · Comment