2022-08-18 Force Field Release Meeting notes

Impromptu Zoom Meeting - August 18
VIEW RECORDING - 49 mins (24 mins of highlights)

@00:01 - Diego Nolasco (Open Force Field Initiative)

So, just to start, a very nice email from Thomas Fox saying that the new version of the annual report addresses his concerns and he's very happy with it.

He said that he is sorry because it's not up to him to change BI's mind with regards to continuing their support to open FF, but at least we know that the annual report now is fine. So good thing to know.

@01:04 - John Chodera

I think you guys have to be proud of yourselves. Right? What this means is that you've done so much good work on the force field side that it's no longer like the burning dumpster fire of the community.

Force fields are at least in a good enough position that people are excited about the direction you folks are taking it in. So I think this is not a loss, it's a win for you folks.

@01:27 - Diego Nolasco (Open Force Field Initiative)

Yeah. I'm pretty sure that people trust the work that we do, and we should be proud. We certainly do.

I'm working in here for a short period of time, and I already feel proud about the work that I am able to do do right now.

And because that feedback about the annual report triggered some discomfort with regards to the roadmap, Looking at the website I saw that in 2022 we were supposed to have Sage 2.2.

But when we look at the improvements, they don't really make that difference because we would have WBO on Stage 2.2, for example, and it turns out that WBO didn't work that well. Right?

So the most important thing I want to discuss today is the fact that we need to come up with a process to address the wants and needs from our industrial partners. I am talking about those gaps that Chris Bailey was talking about yesterday. And I would be really happy if we could start a process to bring the advisory board together while we work on the next roadmap.

What do you guys think about that?

@03:24 - John Chodera

In the past, we had these in person discussions where we got a lot of folks together, including industry, to really talk about a lot of the science, exciting science, and then kind of align the roadmap for what would go to the near term science and then the infrastructure pipeline based upon a lot of feedback.

And everybody kind of fell out of the process because they were able to stay there for a couple of days.

Right. We haven't been able to do that in the last couple of years. So we need to find out a substitute process or a way to do something like that again, to get as you point out, you want everybody kind of feel part of the process so that they are invested.

In the work that's going on in both the science and infrastructure side. We also have to find out how this is going to deliver impact or how we think it's going to deliver the small molecule world, which is what they really care about, small molecular interactions with targets or other targets.

So we just have to figure out what does that process look like and then how do we then work with your team on open source of consortium and then the NIH team to figure out this is research, this is infrastructure and consortium roadmap and then present for everybody as part of this process.

But you have to continually show it to them over and over again so that they really feel bought into it.

@04:43 - Michael Shirts

Now we're going to be getting this and the next year we're going to be getting this.

@04:47 - John Chodera

But do we want to do this virtually or do we want to do this with a strong in person component, is the real big question?

@04:56 - Michael Gilson

I actually think that we could figure out how to do it with a strong in person component in a way that people would feel comfortable.

That would be good because it builds a sense of camaraderie. And I think, John, you're talking about a group that beyond the advisory board.

It's really everybody at the companies who's interested, because I think that it creates its own excitement that then helps people find all they can then convey back.

@05:31 - Michael Shirts

Or it doesn't. If it doesn't, then we better know that. But I think there's two levels here. I think there's the, okay, let's get straight in our mind which of these things we've already been talking about how do they actually fit in?

How can we slot them in to get the next six months, maybe the next year, figured out? And then there's the what's the future of open source field over the next five years?

I think for the first one, we can probably do virtually, and it's probably worth doing it just to get the answers to these questions.

For the longer one. I do think we need to be in person.

@06:03 - John Chodera

Our opportunity for in person is May. So in May, we have an open alchemical free energy meeting in Cambridge, Massachusetts.

So I think we've been tentatively planning to have two days there. The announcement just went around for the free energy meetings.

So we should get our act together and ask people to hold the other day. That would be our opportunity to have a couple of days of in person meetings and discussion and planning with a lot of the folks who are really bought into this.

Let me forward it to you right now.

@06:33 - Michael Shirts

Okay. I'm so out of it that I'm no longer a person who gets invited to open free energy meetings.

@06:42 - Michael Gilson

Is that soon enough? I'm not trying to be difficult.

@06:45 - John Chodera

I mean, it could be that, like, we have to get people to sign on for right now. So we need something immediately that's virtual.

And then we also need to work on the renewals for next year in October, which you're. Yeah. That process really should start in May.

@07:04 - Michael Shirts

Yeah, for NIH. We might need to be going in June. Well, I have to put some budget numbers together, but we might want to apply in June next year.

Not in July, not October or November, where it says it's renewal.

@07:24 - Michael Gilson

I think that the other thing that you made a distinction, Michael, between shorter term concrete decisions and longer term.

I think that had to do the other access, which is consortium stuff versus initiative stuff.

@07:46 - Michael Shirts

I think in the short term, you're talking about what goes on consortium roadmap. Right. Is that right?

@07:52 - Michael Gilson

Right. Or is it also initiative?

@07:54 - Michael Shirts

No, I think the initiative stuff is what are we committed to getting done? Before we write the renewal, like, what do we absolutely have to get done such that we feel like the renewals get our chance of being renewed, for example?

@08:07 - Michael Gilson

I think if we don't have a protein force field that's reasonable, by the time we go for renewal, we're going to have problems and that but on the consortium side, I think that we need to have something workable there that our funders can sell to their management that they yes, correct.

@08:28 - Michael Shirts

I think. What are we going to do over the next six to nine months to get everything lined up for the renewal that can't wait till May?

@08:38 - John Chodera

Yeah. The three tasks are basically, like, immediate. We need to get people to sign on board because they have clear vision of what's coming for open force field next year.

I think you folks have a good road map set already. We just need to get folks excited, and we may also need to do more to find additional people to I'd like to know exactly what that is.

@09:00 - Michael Gilson

Is? Actually, I'm not sure. To me, it's not entirely settled.

@09:03 - John Chodera

Okay. Yeah. So either we need to decide that or but somehow we need to communicate what's coming.

@09:08 - Michael Shirts

I think 70% sure what's coming.

@09:11 - John Chodera

It's not quite nailed down. The next step is we need to figure out what we're doing for the initiative, because we need to rewrite that grant.

And so we need to have an idea of the interesting scientific directions we want to pursue. And then finally, we need to get everybody bought into this process in May and kind of developed a longer term roadmap so that people see how these fit together.

And one of the big questions is, how should this work in general? Should it be that we, the science people, get together and say, here's the six things that we're working on is interesting science projects as part of the NIH funding, and then this is the pallet of stuff that we offer Diego, Jeff and Lily to release.

Here's the possibilities. We can have the industry folks provide input, but eventually you'll get to see what the road map is for the consortium based.

Upon the five things that we're feeding you as possible. Or maybe you just do other things that are not on those because more infrastructure is needed.

So what should that process look like?

@10:11 - Diego Nolasco (Open Force Field Initiative)

Yeah, so we have the fact that we do need two road maps. Right? We need one for the initiative and another one to the consortium.

This is caused by the fact that we are funded by industry from the consortium side and by NIH from the initiative side.

So we have two different things, and we need to take care with compliance because Catherine was kind of asking about that specifically.

Okay. These are the causes. And what are the options? The options, I think they are building a scientific roadmap first and electing the most feasible scientific advancements that we could kind of handle immediately.

And then we bring this to the consortium roadmap, I mean, the infrastructure side of the entire thing, trying to address the so called gaps that Chris Bailey was talking about yesterday with this scientific advancements that we commit to.

Was that an accurate summary of everything?

@11:45 - John Chodera

Yeah, it's just a matter of what's the process by which all of this flow happens. So is it something where science PIs gets to decide that?

There's the six research projects that we'll work on for the NIH thing kind of collectively? Around this core but independent different directions or is it going to be more coordinated in some way with the infrastructure roadmap?

Either can work. It just requires more or less coordination and then if it's all just six independent research directions we can keep updating like here's the six things that we could potentially pull into the consortium roadmap what do you want to hear about the advisory board meetings could basically vote on we want to hear more about this technology and at the next meeting we discuss it and poke some polls in it and then you kind of call them and say what's most important in prioritizing our infrastructure roadmap and then you folks have your direction of quantitative data from the industry partners about what they think is the most important that you can then take into advisement as you set a realistic roadmap.

So a process like that is just one possible concrete way we could do this but very excited to hear other ideas just looking for things that kind of decoupled the different parts in a certain way while keeping people invested and bought in providing.

A workable way for you to do your job too.

@13:02 - Diego Nolasco (Open Force Field Initiative)

And this scientific ideas do they come first? Then we talk to the advisory board asking for their needs?

Or should we first ask what they want or need and then we talk about science?

@13:21 - John Chodera

I mean, you want bi directional, right? But one way to do this is to have a collection of Pis who have their own interesting ideas for how things can develop, right?

If we could just have the initiative fund a collection of research projects that are more forward looking, that build on the infrastructure but don't necessarily slot directly into it, then the palette of options or menu of options can be brought to them.

Which of these things at the current state do you want to hear more about? And they can vote like we want to hear about this at the next meeting.

Tell us more. Give us a whole 40 minutes and talk about X. Or 30 minutes. Talk about it and discuss it, and then, asynchronously you can get feedback whenever you need to in the cycle of the annual cycle, which are the most important science that we should prioritize for building onto the infrastructure map, it's not a constraint.

Right. They're not telling you exactly what to do, but they are giving you quantitative input about how to set your roadmap.

Then when you're together with Lily and Jeff, it's just one model that could work. I don't know. What do you think, Michael?

@14:31 - Michael Gilson

What comes to mind? I think that's a good idea. I think that in terms of two things. One is in terms of giving them the menu of things that they might want to hear more about.

I think that each one should be the menu should have a little list of not just what it is, but why we think it's important and then what the challenges are and some estimate of cost and effort.

So it's sort of like a little bite size.

And then the other thing is just sort of to ask. Do we think that there may be ideas coming from industry that we would not be able to predict that we should let them have?

@15:09 - John Chodera

Are there any things from them that we can we want to give you bi direction, obviously.

@15:15 - Michael Shirts

I think surveying needs earlier rather than later is like, what do you see the needs are? This is a little bit of an overstatement, but I think that they're the ones that know the needs on average.

And on average, we probably can do a better job in brainstorming how to meet those needs, especially since we can see all the needs.

So I think understanding their needs and then thinking about how the science we know we can do is aligned with that, and then presenting but really, it has to be featured.

Presenting a collection of options, getting feedback on that, and iterating. I think we should get to know we should know needs first.

@15:53 - Michael Gilson

I think that's a good way of doing it. It's something that we can tell them. I don't know if Diego might be the point person.

I don't know what it's say. We're in a planning exercise and we've accomplished a lot. We're in a planning exercise.

We'd like to know more about your needs and go out to it and where do we bring your ideas and coming back to you?

@16:11 - Diego Nolasco (Open Force Field Initiative)

At OpenFE I generated a form that people would use to propose new items to our backlog.

To be honest, I was hoping that they would propose features, but they were starting to be too specific because they understand some science about what we are doing.

Right. But we can come up with the same thing so that they could feel engaged, they could feel included from the beginning.

And so I could kind of send them this one form. They could write down their expectations, and their ideas for features.

They could describe some science they would like to see on the force field, and then we could start this bidirectional thing that John was talking about.

@17:16 - John Chodera

Does that work to capture a lot of input, or is this not implemented yet?

@17:21 - Diego Nolasco (Open Force Field Initiative)

It is implemented. Clara Christie, I think she's from Buyer I'm not sure yet, she gave a bunch of ideas.

She got, like, crazy.

@17:34 - John Chodera

It's a great mechanism if we can get other people to use it too. I do wonder if we need to be a bit more active, though.

So if we just have it passive like that, then there's other folks that might not really contribute. Even if so, they run into challenges a lot.

So I do wonder if there's another option to have every advisory meeting, one of the representatives is randomly chosen ahead of time to say, tell us five minutes about a spectacular failure here.

To run into this back here, of force field and just to understand different modes of failure in a way that they can actually talk about.

Getting kind of poking people to give you information or examples of what's not working is really helpful and inspiring new science even if the science isn't.

We don't even understand it yet.

@18:22 - Diego Nolasco (Open Force Field Initiative)

That is a good idea. I am afraid of putting people in bad situations if they are kind of not willing to say too much.

But that's a very good idea. We could do both, I think. Or even come up with another process. We need multiple mechanisms.

Okay?

@18:43 - Michael Gilson

Basically it goes meetings, but everybody gets a plate. Person gets a platform. They could talk about a spectacular failure or about an idea they thought it could be.

Just bring something interesting.

@19:01 - Diego Nolasco (Open Force Field Initiative)

Okay.

@19:02 - John Chodera

Any other fairly type of insurance like that that would be ideal. Rather than volunteer, you already have a mechanism for people to volunteer information.

@19:15 - Diego Nolasco (Open Force Field Initiative)

Can you say it again please, John?

@19:16 - John Chodera

So you already have a mechanism for people to volunteer information. So for something like this, it's better if everybody has to take their turn in some way so that it's fair.

@19:28 - Diego Nolasco (Open Force Field Initiative)

Okay, I'll think about a way. Do we have any other option? Any other idea about how to do that?

@19:46 - John Chodera

I would just put a roster together and have a roster where you just designate someone and we say the next month, next we'll talk about something from a new idea or a challenge that they run into.

@20:01 - Diego Nolasco (Open Force Field Initiative)

So that they could prepare themselves to do so?

@20:03 - John Chodera

Yeah. Okay. And keep it easy, informal 5 minutes. Show a slide or two if they want.

@20:09 - Michael Gilson

Exactly.

@20:10 - John Chodera

Yeah.

@20:12 - Diego Nolasco (Open Force Field Initiative)

Okay. Yeah. I think they will like it, to be honest, because then they will have the chance to show their needs properly and we will not need to keep guessing. Because this is how we kind of do today.

Right?

@20:34 - John Chodera

The industry benchmarking season was probably the most important part of it was like getting feedback about how things actually work inside of industry domains.

Right. Like making sure you lead people through this exercise. Everybody has to share something. You have to run it.

They have to share things. Finally you get all of their opinions about how things should work and all that feedback back and that's very valuable.

@21:01 - Michael Gilson

Yeah. And I think that by getting the quiet people to speak, it's a little bit alternative to a workshop.

@21:11 - Diego Nolasco (Open Force Field Initiative)

Okay, so now we have options. I think that the action is all about me, right? I should kind of delineate what we have just said now, and then I will submit it to your approval.

@21:36 - John Chodera

This was just one possible mechanism, though, so it would be useful if other folks have other ideas. We've articulated where we basically have a collection of science projects in the grant that flows.

So here's the panel of options you can hear about as industry folks. What do you want to hear about next month?

And then you hear about stuff in more detail and discuss it. And then at some point, you survey folks to get quantitative information.

Which science bits are most exciting to them maybe they get 2nd 3rd. Whatever you get some quantitative information that helps set your roadmap with all this community engagement and numerical reports and then we have some mechanisms to take what science should be focus on that goes all the way back right so that's just one possible model but are there other better models we can do things that are more coordinated?

@22:25 - Diego Nolasco (Open Force Field Initiative)

Well, I used to use a tool named Design Thinking and I don't know if you guys are aware of what design Thinking is. It’s a very nice tool especially when we are together so that we can kind of point out what are the feasible possibilities on each other's ideas and then we start brainstorming on fixes for those ideas. I should come up with a way to do it virtually. I never did it before.

@23:05 - John Chodera

Alternatively, it could be an action for an extra day at the workshop. Right. We could do an exercise like that for helping set our road map or flushing out ideas roadmap in the in person one, because it does work really well.

@23:21 - Michael Gilson

Going back to the advisory board thing, I think, John, you were saying about having each meeting a different person would do it, but another way to do it would be to make it a micro workshop by having one meeting where multiple people I don't know how many we could fit would get up and testify and have a little discussion.

I think that might be more fun and just you a whole bunch of ideas and things might click better.

@23:50 - John Chodera

That would certainly be very useful. Now, thinking about what are we going to put in the grant?

@23:57 - Michael Shirts

I think it might be easier to get a bunch of people to get together for limit it to an hour and a half or something.

Or even an hour or something.

@24:07 - Michael Gilson

Rather than committing to short chunks of time over a long period, how many people would we be wanting to have safety works?

@24:19 - Diego Nolasco (Open Force Field Initiative)

Well, I think we have around let me just check. How many people do we have on the advisory board?

Do you guys remember that?

@24:33 - John Chodera

I think we have more than a dozen. But would we be able to get through six in a session, do you think?

@24:39 - Michael Gilson

Five minutes each, plus some discussion. If we have a chair who is willing to I think we could get through six in an hour and a half up with that.

@24:51 - John Chodera

It being totally exhausted.

@24:53 - Michael Gilson

I think we just have to at certain point, somebody has to say, Well, there's a lot to talk about that.

Let's hold on to that and move on to the next one.

@25:04 - John Chodera

Yeah, there's like 14 a or something like that.

@25:19 - Diego Nolasco (Open Force Field Initiative)

Yes. If we could get around six people, I think it would be already okay, but the more the better.

@25:45 - John Chodera

My goal would be we could do two sessions each, and that would basically get through everyone.

@25:53 - Diego Nolasco (Open Force Field Initiative)

My goal is to make people feel included.

@25:57 - Michael Gilson

I agree. I think like Johnson. We have two sessions so everybody would have a chance. Something like that.

@26:07 - Diego Nolasco (Open Force Field Initiative)

Okay, so the idea is come up with a kind of a mini workshop so that people could tell about their big failures during the last year.

@26:19 - Michael Gilson

Well, failures or priorities? Needs. It could be a technical idea. I really think we need X to make our models work.

Well, here's why.

@26:33 - Michael Shirts

Although I do think we're better off telling what's not working. Like what should we do? Or what are the needs?

It might turn into that, but I think the emphasis is I think we'll learn more from saying this is failing.

@26:45 - Diego Nolasco (Open Force Field Initiative)

We want to do this and it won't work. Then I think you should add this. Yeah, we could kind of ask so that people bring anything they want.

What didn't work? What are they willing to have? What are their dreams? And then we will have a bunch of stories.

And we should see that as user stories. Right? Because I think we need to have user stories so that what we are developing adds value to the industry partners right away.

@27:29 - Michael Shirts

No, I think it works everywhere. I think it works like the NIH as well, because we're doing this because companies need it.

@27:36 - Diego Nolasco (Open Force Field Initiative)

Okay.

@27:37 - Michael Shirts

That can also be more of a justification of things seem more fundamental. It's like, yeah, we could do something sexy.

But one of the things we're doing is this because it's absolutely needed.

@27:51 - Diego Nolasco (Open Force Field Initiative)

Right.

@27:52 - John Chodera

And we have companies that say, so do you think we could do this, like, next month in the advisory board slot and just.

Reserve 3 hours instead.

@28:05 - Diego Nolasco (Open Force Field Initiative)

I can check with people if they are able to stay longer because it seems like a big ask.

@28:14 - John Chodera

We can do two months, an hour and a half each.

@28:22 - Diego Nolasco (Open Force Field Initiative)

Yeah. And maybe we will not even need an hour and a half because maybe people tend to be shorter and speak less, I don't know.

Or will that trigger discussion from our part? Is that the idea?

@28:43 - John Chodera

So I think if everybody presents about five minutes and we have about five minutes of discussion, in principle that fills an hour, but I'm guessing it will go a little bit over.

Yeah.

@28:54 - Diego Nolasco (Open Force Field Initiative)

So I can certainly check if people are willing to devote their time for the next month board meeting and the other one.

Right. September and October.

@29:08 - Michael Gilson

Which date of those?

@29:17 - Diego Nolasco (Open Force Field Initiative)

So it will be September 21 first and October 19. Off topic question when is your daylight saving time ending?

@29:51 - Michael Shirts

First week of November.

@29:57 - Diego Nolasco (Open Force Field Initiative)

Thank you.

@29:57 - Michael Shirts

Let me get the exact date. November. November 6.

@30:01 - Diego Nolasco (Open Force Field Initiative)

Okay.

@30:04 - Michael Shirts

Around the end of October. There's always some weird things. Europe changes earlier, doesn't it?

@30:15 - Diego Nolasco (Open Force Field Initiative)

I think ours here in Brazil is about to start in November.

@30:20 - Michael Shirts

Yeah. Southern Hemisphere.

@30:22 - Diego Nolasco (Open Force Field Initiative)

Carmen is about to switch to being closer to our time or further. I was just checking. Yeah. It will improve my life a little bit.

We have action now. We have action to take.

@30:42 - John Chodera

Does anyone want to share any other idea? Yeah, just to come back to this last question, though, michael, any other ideas about how we could do how the initiative should feed into the consortium in terms of should we work together more closely or this idea of.

We kind of do what we want and then flow them, give them a menu of options.

@31:06 - Michael Shirts

I think consortium and initial need to work together a little more closely in terms of the leadership. I think it should work together more closely.

Things like working on surrogate modeling that'll probably be brought in, but eventually but I think just what scientific ideas you're going to explore is going to be affected by what tools you have.

Right? Yeah, I think working closer together is better now, again, then you get the point where these priorities are this and these priorities this.

Okay. We have to diverge at that point, but I think the closer it works together, the better the rate to make sure the divisions are clear dividing who gets funded by what, by what sorts of money, but that the project should be.

Integrated to the extent that it makes sense. Like, the default should be work together more, but they'll be imperatives that drive it apart.

@32:12 - Michael Gilson

So my media thought is, I feel like we're in a moment where we need to work together because we're sort of figuring out where we are and what's what these are and what we can do.

So now, definitely. But going forward, let's say once we set it here and room out there, they may not need to interact that closely.

Right. For example, like for the protein force field, do we want input from industry as we develop that? Maybe I could use an explanation of what you mean by working more closely once we get past a new planning phase.

@32:55 - Michael Shirts

Yeah, well, one way is that we want people doing the new science, the new ideas, to the extent possible to be using open source, using existing technology, because that'll make it, one, it'll make it easier automated to it will make it way easier to bring stuff in.

And so there should be some discussion. And if there's choices about design, obviously the priorities, what industry wants, has to dominate.

But there's usually like three different ways to do things and one of them might make it really easy to do some of the science and two won't.

@33:29 - Michael Gilson

That's a given.

@33:30 - Michael Shirts

Michael I think we all want to work with the core infrastructure and then ideally try to slot it into groups of people that don't talk, that talk less you start getting drifts.

@33:46 - John Chodera

Here's is what I mean, though. How do we set the set of so we know that Diego is driving, along with Jeff and Lily, sending the infrastructure roadmap for the consortium?

@33:59 - Michael Shirts

Everything.

@34:01 - John Chodera

It for the initiative. For the NIH funded side, do we jointly develop with the consortium representatives exactly what research projects we put into the grant work on, or do we choose that independently and then present a menu of options to take into the consortium?

I think this is the real question about scientific directions. How do we select those? Do we have to do integrated scientific roadmap planning or can we decouple that?

It's basically the main choice.

@34:30 - Michael Shirts

Yeah. I think for the NIH science, we should get input from industry, so we should get input from the consortium of what's needed because that will just make it stronger, it will motivate it.

But I think that the science roadmap is affected by the infrastructure roadmap because what infrastructure is going to be available and they interrelate. So, I mean, is it the same roadmap? I don't know, but it seems like. I don't know.

It seems like it should be.

@35:18 - mattthompson

We can have different roadmaps that are highly coupled.

@35:21 - Michael Shirts

Yes, that's what I would imagine. Different roadmaps that interrelate, that talk to each other. They are aware of each other.

Maybe that's the way they are, separate roadmaps, but they are aware of what's happening in the other roadmap.

@35:37 - Michael Gilson

What do you mean by the road? Are there dependents? I think maybe one of the things we criteria for something going on the Consortium roadmap is that it's not dependent on something on the science roadmap.

Right. The science is doing that. They turn out this way, it makes it out that way, but it's not going to take it's not dependent.

@35:57 - Michael Shirts

But on the other hand, if you say: what is the consortium going to be implementing three years from now?

Well, odds are it's going to be taking the science roadmap, or there will be things developed in the science roadmap that will be that right.

Because that's the whole point.

@36:11 - Michael Gilson

The whole point of the science has developed stuff that can go in really independent because but those aren't on the road map yet.

Right. The things that are on the road map are ones that ones that are going to be on the road map, they're not there yet, but they can come over.

@36:28 - Michael Shirts

Yeah. It depends on the five year road map.

@36:30 - Michael Gilson

Yes.

@36:33 - Michael Shirts

Is it a one year roadmap or is it a five year roadmap?

@36:37 - Michael Gilson

I guess the thing, though, is also, even if the five year roadmap or maybe the question of what the things on the roadmap are yes.

@36:52 - Michael Shirts

What we want is we don't want the consortium to be held back from making a release because of science.

But it's still going to be science that's flowing into. The infrastructure.

@37:04 - Diego Nolasco (Open Force Field Initiative)

One thing we should take care is that Katarina is afraid of facing some compliance problems if we fund long term science with industry money.

@37:30 - Michael Shirts

So I think this kind of let's be specific, NIH compliance company compliance, let's be precise. I don't have the answer for you, I think. Okay. Because there's certainly compliance issues if we are getting paid twice to do the same thing.

@37:37 - Diego Nolasco (Open Force Field Initiative)

Yeah, I think she's afraid of that.

@37:40 - Michael Shirts

Yeah, well, that I think we should and that we probably should have been worrying about a little bit more than we have.

But that's the only complaint compliance issue I can think of. And I think this is going to be reflected in the NIH proposal.

I think open source does counted. Kind contributions as generally. We'll have to figure that out. But I think that's the only compliance thing that I can think of, so which can be addressed by making it clear which long term science NIH is funding.

@38:15 - Diego Nolasco (Open Force Field Initiative)

Yeah. And I think it makes mandatory that we have two roadmaps, two separate roadmaps that of course need to be aware of each other.

@38:25 - Michael Gilson

You're saying each roadmap corresponds to a funding stream?

@38:29 - Diego Nolasco (Open Force Field Initiative)

Kind of.

@38:31 - Michael Shirts

That might be a way to think about it.

@38:35 - Michael Gilson

Makes it a lot easier to say, okay, this is what you're paying for. Yeah, that's true.

@38:43 - Michael Shirts

Yes. Somebody asks.

@38:48 - Michael Gilson

Yes.

@38:51 - Diego Nolasco (Open Force Field Initiative)

And of course, this roadmap, they need to talk to each other and we can do it kind of internally.

We are doing that right now by using the WBSes because in there we have infrastructure and science put together, and people don't really need to take that because that's technical stuff.

So it's all about us. People need to pay attention to the features, how we are developing each of the features.

It's on us. So we don't need to give details on that. But, yes, I think we should have two different roadmaps.

How are we starting the process? Because John said, oh, I can fly to San Diego. That's not that easy for most of the people involved.

I think that, for example, if you you talk to the industry people, oh, can you fly to San Diego next week?

@39:59 - John Chodera

If we're going along the lines of what I suggested here. The way to start would be your industry challenges and things that don't work that will help provide a lot of material for us to look at in terms of meeting along with the PIs maybe also putting you folks from the consortium as well.

It would be great to hear your input but then we kind of plan all the research directions that we want to prioritize in the NIH part and then come up with what we think we're going to pitch there.

Maybe we also run that by the industry folks for feedback. We then put that feedback right into the ground and it would make it very easy for us to write.

We plan on an in person meeting in May where we have all of the industry and initiative folks together where we do some joint roadmap for the longer term clients to talk about other things.

But I think then we try to flip this idea of we'll give you a menu of options of here's research projects in various stages.

Choose which one you want to hear about and. Advisory board meeting talk on it and get your feedback in your point you think it's ready to when you need to do your infrastructure prioritization you can get quantitative feedback about the menu of options that we can put them on website Mike suggested have a little blurb about each one like a mini presentation type thing about what we think the resourcing is et cetera.

To get started, 1) schedule those industry feedback sessions then 2) we do some infrastructure planning, 3) we have them reserved the May slot for in person meetings for long term roadmap and science.

I think those are the critical things.

@41:45 - Michael Gilson

Just adding to that I think that the sort of near term advisory board presentations of ideas and needs I think that can feed into both roadmaps because some of those are going to look short term.

Oh, we really need virtual sites bam okay we got that. Some of them are going to be long term. We need to integrate AI somehow. Okay, well, there's a bunch of ways of doing that that's long term. Yeah, I think so.

@42:00 - Diego Nolasco (Open Force Field Initiative)

Okay, so action now is coordinating all of this. I think it would not be that hard to bring people together and make them speak for a little bit.

@42:39 - Michael Gilson

I think they are willing to do that, given the sort of reach of what we're talking about right now.

It'd be good to see what David thinks before maybe executing. Diego, would it make sense for you to write down, like, a paragraph about what you think we decided, and then among us, including David and other people?

@43:12 - Diego Nolasco (Open Force Field Initiative)

Yeah, sure. And we also have the meeting on Monday, and David is invited to participate on that. Jeff, Lily and I have the lead team meeting on Mondays, Mondays and Wednesdays, and David is invited to participate in the next one.

So he will be there, but I can certainly prepare something and share it on the appropriate channel and ask for feedback.

And, of course, maybe maybe I will not write it the proper way and you will be able to write together.

We can make sure we're all almost literally on the same page. Yeah. Yeah. So we have a kind of a decision about the action that should be taken.

Do we all agree on that?

@44:17 - Michael Shirts

Well, it's is what we're going to do after we get the input. What's the process? After the input, the governing board has a meeting and actually what yeah, after we have these meetings, what's the next process?

@44:35 - Diego Nolasco (Open Force Field Initiative)

Well, I think that we should treat the inputs as user stories. We should evaluate all of them and give feedback.

Of course. And among us, we should prioritize. I think there needs to be different levels of prioritization. Because we have different things running at the same time but PIs need to prioritize scientific development and the tech team needs to prioritize software development and by doing so I think we will be on a very good path to put things together and come up with technical solutions to the problems that they bring to us both afterwards in the realm of possible.

@45:38 - John Chodera

So we need to have a scientific solution we can explore to address the story some of it is probably not going to be we lost the final part of what you were saying that we also have to work in the realm of.

And have to address the things that address all. Even if it's like can you make it have zero error?

Yeah, maybe you need to specify realistic.

@46:21 - Diego Nolasco (Open Force Field Initiative)

Hopefully we will be able to say no to some things because I am very worried about scope creep because sometimes people just want too much and it makes projects fail at the beginning.

So I know that my job is kind of being the boring person at the moment, but, you know, you guys are paying me to do that.

But if we do it together and if we come up with strong arguments about why we are not prioritizing this, and we are prioritizing that and make them understand in a technical and scientific way, I think everybody will be happy at the end.

@47:18 - Michael Gilson

Yeah. Including us. If we have 12 or 14 people give their ideas, I don't think there are going to be 14 really different ideas without overlap that have to be seriously considered.

I think there'll be a fair amount of overlap.

@47:40 - Diego Nolasco (Open Force Field Initiative)

I think it would be just like saying right now people will say, oh, I want to have virtual sites and that's it.

Because people are very hyped for virtual sites right now and we have it implemented in the force field. It can handle virtual sites, but we have to generate data. Because of that, we can't attach the release of a force field to the development of virtual sites. It's not that easy. People tend to think that it's really easy.

Like, I want to have it, the developers do it, and then it's done. Here it is. But there's a whole process, and we need to respect that if we want to be recognized as a force field that generates reproducible data.

Right. It's our name. Okay. Any other comment or question about everything we discussed? I think that we just had a very fruitful discussion, to be honest.

@48:53 - Michael Shirts

Sounds good.

@48:54 - Diego Nolasco (Open Force Field Initiative)

Yeah. Okay, so thanks, everyone. One for showing up. I have a bunch of work to do with the recording.

With the transcript. And I promise to share the transcripts with you as soon as I have them ready.

@49:14 - Michael Shirts

Sounds good.

@49:16 - Diego Nolasco (Open Force Field Initiative)

Thank you.

@49:18 - Michael Gilson

Bye, everybody.

@49:19 - Michael Shirts

Bye.