Hello everyone:
First of all, excuse these days without notices from us, but the start of the course has required more time than we expected and we have not been able to be as active as we would like in the project.

We update you on how the optimization of the model is going: after 3 iterations of the algorithm, the model is gradually improving, but one of the markers is greatly limiting the improvement. We full understand why this is happening, so we need to stop and analyze it in detail before continuing.

Thanks to all the simulations you have done we have a wide variety of situations to study and we hope to resume the simulations shortly (ideally 1 week, but it is easy for it to be two).

On the other hand, the applications that we had in Beta mode have already been sufficiently analyzed to be considered stable, and until we modify the application again, the beta tasks will be stopped.

Thank you very much and we will keep you informed shortly.

All the best,
Jesus.

=================================================================

Hola a todos:
En primer lugar, disculpad estos días de parón sin notificar, pero el arranque de curso nos ha requerido más tiempo del que esperábamos y no hemos podido estar todo lo activos que querríamos en el proyecto.

Os ponemos al día de cómo va la optimización del modelo: tras 3 iteraciones del algoritmo, el modelo va poco a poco mejorando, pero uno de los marcadores está limitando muchísimo la mejoría. No tenemos del todo claro por qué está sucediendo, pero necesitamos para y analizarlo en detalle antes de continuar.

Gracias a todas las simulaciones que habéis hecho hasta el momento tenemos gran variedad de situaciones para estudiar y esperamos reanudar las simulaciones en breve (idealmente 1 semana, pero es fácil que sean dos).

Por otro lado, las aplicaciones que teníamos en modo Beta están ya suficientemente analizadas como para considerarlas estables, y hasta que no modifiquemos de nuevo la aplicación, las tareas beta estarán paradas.

Muchas gracias y seguiremos informando en breve.

Un saludo,
Jesús.

---

Jesús Carro
Universidad San Jorge
@InSilicoHeart

There is a mistranslation of"No tenemos del todo claro por qué está sucediendo,"
"We full understand why this is happening" reads like you already know whu this is happening.

There is a mistranslation of"No tenemos del todo claro por qué está sucediendo,"
"We full understand why this is happening" reads like you already know whu this is happening.

Google translates as:
"We're not entirely clear why this is happening,"
That is the opposite of original.

Paul.

Any update as to when this project will resume?

Hi Greg_BE,
We are still analyzing the problem. At least one week more.

Best,
Jesús

---

Jesús Carro
Universidad San Jorge
@InSilicoHeart

Hi Greg_BE,
We are still analyzing the problem. At least one week more.

Best,
Jesús

Thank you for the update!

Another update at 17.10.2022?

Jesús,

What is the status of your project this week?
You had a good start and then suddenly stopped.

We do not fully understand what the 'problem' is that you have found in the work we have returned. Can you explain what the 'problem' is and if you are close to a solution?

I think for now I will put your project on 'hold' and check back at the end of the month. Or maybe you get do a broadcast notice that shows up in BOINC so we know your running again?

Hi all,
It is taking us a little longer than expected to analyze the problem.

It is not easy to explain it, but broadly speaking it would be the following:

1. To improve the model there are a series of parameters that we can modify.
2. The simulations that we send to your computers replicate experiments that have been carried out on human ventricle cells and that have been published.
3. Every simulation runs a different protocol with a different combination of parameters.
4. With the results that you send us, we analyze them to calculate the same markers that were calculated in the laboratory with the experiment you have replicated.
5. With this we obtain about 100 markers for each combination of parameters that we can compare with the experimental results.
6. For each combination, it is scored with the differences between its results in the model and the experimental values.
7. We have an optimization algorithm that tries to find the combination of parameters that gives the best score (which would be the closest to what has been seen in the experiments).

There is one of those 100 markers that is not getting anything like what has been seeing in the experiments and no matter what we modify we do not get it. We need to check carefully what it happening. It could be:

a) A limitation of the model that is not able to reproduce it (in this case we will probably leave this marker temporaly out and address it in the next step of the investigation)
b) Some kind of error in the calculation of this marker in step 4.

Another limitation at this moment is also time, finding this problem has been combined with a moment in which we have little time available to dedicate to this project (we have others underway and at this moment, this one should advance a little more autonomously). I'm the first who wants to clear up this question (I can't get it out of my mind), but we'll have to delay a couple more weeks.

I don't know if the explanation is good enough, or if it was understood. I can give more details but I don't want to lose you among more complex things. If you want to known more, ask and we'll continue talking about the marker.

Thank you for your patience and for your interest.

Best,
Jesús

---

Jesús Carro
Universidad San Jorge
@InSilicoHeart

Thank You Jesus,

So your saying essentially is one of the data points is not conforming to the science.
Maybe it would be faster to just drop that problem and send us the rest of the data.

Is this your only distributed computing project?
Do you have something else you can send out to keep the project going while you figure out what's going on with the current project?

I just suspended your project for now and will check back end of the month.
But a project with a month of no work, people may leave and may never come back.

But a project with a month of no work, people may leave and may never come back.

People will come back fast enough if they know about it.

Be sure to post a notice on the BOINC forum.
https://boinc.berkeley.edu/forum_index.php
The "Projects" section should do it.

Hi Greg,
I answer your last questions:

So your saying essentially is one of the data points is not conforming to the science.

The result of the model is not conforming to the experiments. No model is perfect, we are always trying to get better models, but there will be always things to improve.

Maybe it would be faster to just drop that problem and send us the rest of the data.

The process is iterative: we send you 1000 combinations of parameters, and using this information we try a best set of 1000 combinations of parameters. The decission of removing one experiment affects to the next combinations.

Is this your only distributed computing project?

Yes, this is not the only research we do, but we have only one BOINC project.

Do you have something else you can send out to keep the project going while you figure out what's going on with the current project?

No yet, we have some ideas but we need to prepare them.

I just suspended your project for now and will check back end of the month.
But a project with a month of no work, people may leave and may never come back

Possibly, we are aware, but we cannot work faster. As has been said in another message, we trust that you will come back and as soon as we are going to launch simulations again we will also notify you from the client.

Thank You Jesus

Thank you!

---

Jesús Carro
Universidad San Jorge
@InSilicoHeart

Jesus,

As your tech team how to do a "broadcast message" via BOINC manager notices that reflects major announcements like starting up again or server problems, etc.

When you start up again, just send out a message that way (via notices) and it will reflect what you put on your news section on your front page.
Then all of will know when your back to work again.

Meanwhile good luck with everything.

Jesus, Thanks for the thorough update and explanation of what you're doing. I'm sure that I'm just one of many BOINCers that will gladly return as soon as you start posting WUs again.
Don't let any worry of lost support cloud your best scientific judgment.

---

I'm sure that I'm just one of many BOINCers that will gladly return as soon as you start posting WUs again. Don't let any worry of lost support cloud your best scientific judgment.

Agreed. The unfortunate reality is that every project I've been part of has people who start loudly telling the forums that they're leaving, anytime there's a pause in work. Don't worry about it.

Also, just in case anyone reading this isn't aware, you don't have to leave a project at all. BOINC is perfectly happy to let you attach to multiple projects at the same time, and lets you balance their priority. When one project doesn't have WUs, BOINC will just work on the projects that do. I'm currently attached to DENIS, Milkyway, and WCG, for instance.

Ready to rock whenever you are, keep up the good work

SV

No news good news !
See you later
K.