Ruth Petersen

Welcome to another installment of Voices of Sample Management. I'm Ruth Petersen. And today, I'm joined by Derrick Miyao, Vice President of Molecular Foundry at DeepCure.

Before DeepCure, Derrick spent 20 years at Neurocrine, where he made significant contributions to automated drug discovery. Notably, he engineered 2 systems that can synthesize up to 10,000 compounds a day, showcasing his unparalleled expertise in sample management.

Thank you for joining me today. Derrick, can you get us started by telling us about the work that you're doing at DeepCure?

Derrick Miyao

Yeah. Hey, Ruth, it's a pleasure to be here today.

So at DeepCure, we're not too far off of what I did at Neurocrine except for the fact that we're focused on small molecules using an AI platform using modern technology to help identify novel treatments for immunology or inflammatory diseases.

And what's unique about us is we've gone all in on our platforms whether it's AI or a molecular foundry, we truly believe that these technologies will get us to drugs faster.

Ruth Petersen

That's wonderful.

Let's see. So how are you using Mosaic at DeepCure?

Derrick Miyao

Yeah. So Mosaic's a critical piece of our day-to-day functions inside the lab. As you know, sample management is critical to any organization, but especially when it comes to automated chemistry.

The one thing that's unique about automated chemistry is we are using a lot of different reagents that are at different that are that are stored differently, whether they're solids, they're neat, they're in solution.

And so sample management is essential to what we do, especially when it comes to robotics because if your inventories are off, if they're, if your bottles aren't labelled, it's really hard to be efficient when it comes to automated chemistry.

So in that regard, we use in our day-to-day functions whenever it comes to lab functions, whenever we're setting up molecules and synthesis, it's part of our everyday function.

Ruth Petersen

And so and DeepCure is based in Israel, so are all your compounds being stored in Israel or do you have them in multiple locations?

Derrick Miyao

So that's a great question. Right, our headquarters are actually in Boston, but that's just an office site.

Our first lab is in Israel, It's based in Rehovot, which is just outside of Tel Aviv. Yes, there we do store chemicals, we store cell lines, we store final molecules. And then we have a second lab here in San Diego where we are only focused on chemistry and we use it here as well.

Ruth Petersen

Ruth Petersen

So how does what you're doing at DeepCure with regards to sample management differ from what you were doing at Neurocrine?

Derrick Miyao

The only difference in what we do here at DeepCure when it comes to the actual work is the amount of work at a start-up is much less than Neurocrine.

Neurocrine is all about scale, performing thousands if not millions of things per day versus you know hundreds and thousands of things at a start-up.

This scale is just very different when you're at a start-up.

And, so I would argue though it's equally as important in both settings, because in start-ups it's really easy to treat logistics as not as important as the science, which in general generally is that way.

But if your logistics aren't optimal, then your day-to-day life will be difficult and it's no different than when you're at scale performing you know thousands of transfers or dose responses, when it comes to samples, if your systems are not working properly or if the users aren't keeping things up to date, your day-to-day life just becomes far more difficult.

Ruth Petersen

So basically, I mean what you're doing at DeepCure is, in terms of scale is very different from Neurocrine.

But if both of those organizations you've onboarded Mosaic. Why Mosaic?

Derrick Miyao

So that's a great question. With that being said, there's there's quite a few options for sample management.I've seen options from complete LIMS systems where sample management is integrated to it, as well as standalone systems like Mosaic.

And more recently I forgot some of the names of companies that have struggled trying to do sample management.

But with that being said, sample management has been the gold standard in sample management for research groups for the past 20 years.

With that being said, it's not without its flaws, but it does work.

It works well, it has the foundation of sample management and it has 20 years of experience of building out a great product.

And so with that, it's reliable, it's robust and there's also great support within Mosaic or within the Titian team.

Ruth Petersen

That's awesome.

Getting back to to sample management groups, why do compound management groups take all the blame for bad data?

Derrick Miyao

That's a great question.

So, hopefully I don't make a lot of people angry with my answer, but sample management is easy to blame.

They're in the middle. They're not directly responsible for performing the science, but they're very easy to blame.

But with that being said, science is difficult as it is. Performing assays is difficult. They're sensitive. They're expensive.

You know, it's as simple as maybe your liquid handler channel went off and your data's changed and a scientist may have performed an assay 100 times and then that machine has never fared, and so their natural instincts go back to, well, if it wasn't my assay because I've done this 100 times, maybe it was someone who handled the sample in between me.

And so that's typically why they like to blame compound managers.

But with that being said, the best compound managers are the ones that you never hear anything about, and those are the ones you never really praise.

And you typically only find mistakes.

And with it, when it comes to sample management, it is very easy to make mistakes.

It's a job that is not as appreciated as the science. It's a job that is routine and repetitive. So a lot of people can lose focus and make easy mistakes.

And once you make certain mistakes within compound management and you lose the trust of the scientists, they're always going to second guess your work.So with that being said, science is hard.

It's really hard to pinpoint where the problems are, but it's really easy to pinpoint sample managers for being the problem and it's pretty universal.

No matter who you talk to in sample management, they all have stories about how they're the problem.

Ruth Petersen

Do you have any advice for someone who is dealing with, they're constantly getting blamed, kind of, how do you triage the troubleshooting or kind of work as a team?

Derrick Miyao

My advice is, when people have problems, you should listen.You should try and understand them and simply provide the data to back up your your side of the story.

And so at the end of the day, we're all here to discover drugs. We're here to help people.

We're not here to tear each other down. And so my advice is what we do is difficult.

View it as difficult and you're gonna run into some serious problems and so you have to work through them.

And it's no matter whether it's in life or in work settings, just treat people with a golden rule and usually things work out.

Ruth Petersen

Yeah. Yeah. That's wonderful. Thank you.Can you give any advice on home grown solutions versus off the shelf with bespoke touches?

A lot of what we do right is, is never off the shelf, there's always some customization, but that versus a completely home grown solution. 

Derrick Miyao

Is this a broad question in terms of any software or sample management software?

Ruth Petersen

I don't know. I'm trying to think if that applies to both scheduling. I mean, let's start with sample management, but it could apply to scheduling software as well.

I've seen all kinds, even hardware, right, people building systems from scratch on their own and trying to try to piece it together.

So however you feel like you like to answer it.

Derrick Miyao

I think this is a great question and I think it's a question many companies ask themselves every day that they face their software issues, buy versus build.

For me, when it comes to the core of the business, sample management is not core to my business.

So I don't necessarily want to distract my team with focusing on sample management and building software.

One, it's very technical.

Two, it's not as easy as you think it is because there's been a lot of players in the past that have tried to build new systems and really they promise you everything under the sun, but paying for developer time for your customizations that only work for your company.

2nd to that, when it comes to off the shelf, there's no real solution that fits for everybody.

And so even when it comes to off the shelf, you have to modify it for your particular workflows or your particular company.

And, so what I can say is both answers are right and both answers are wrong.

But really, it really comes down to what's around you, what expertise you have, what potential vendors you could use, what internal teams you have.In terms of mine regard, I use both options.

So when it comes to sample management, that's something I'd buy.

And if I have problems, I can simply tell you at Titian, “hey, these are my problems, do you want to work on them?”

And if you say yes, then it's a partnership where we can work back and forth. If you say no, then it's then I have to go and see, Well, I need a different solution.

Maybe it's custom, maybe it's something else off the shelf.

But when it comes to sample management, because it's not core to my business, I really do not want my team focused on redeveloping something that someone already has.

Ruth Petersen

Yeah, no, that makes a lot of sense.

When you're limited resources, you really have to focus them on what everyone's good at.

Let's see, I've got a couple of rapid fire questions for you.

So, first one, automation platforms, large automation systems like more than two robots or smaller work cells?

Derrick Miyao

If I have to pick one, I will say smaller work cells because they're more modular.

But I'm not opposed to big systems when your little systems work properly and you can integrate them together.

But with that being said, I always like to start small and then go big, never start big.

Ruth Petersen

Yeah that makes sense.

Well, keeping that in mind with the smaller work cells, what's your opinion on mobile robots?

Derrick Miyao

I think there's many answers to how we will be performing transfers in the lab, and mobile robots certainly have their place.

We've seen them in homes where Roombas were just toys 15 years ago, but now you see them in every other house, especially in houses with animals.

When it comes to sample management, samples need to be transferred across the lab, cross transferred between stations all the time, whether it's an arm or a rail or robotic mobile robot, it's going to be some sort of version of that that will that will transfer compounds.

And so I certainly believe there's a mobile robot in the future for future labs.

Ruth Petersen

Repository compounds - solid or liquid?

Derrick Miyao

Liquid

Ruth Petersen

Vials or plates?

Derrick Miyao

Vials

Ruth Petersen

DMSO percentage?

Derrick Miyao 

Saturated with water

Ruth Petersen

Sample storage temperature?

Derrick Miyao

Room temperature.

Ruth Petersen

All right, awesome.

And then finally, any words of wisdom for sample managers just starting out on their sample management journey?

Derrick Miyao

Absolutely.

So, when it comes to sample managers, one: you have to understand that you're going to be in a service role, but it doesn't mean you can't contribute to improving the lives of the co-workers that you have.

Also keep in mind that sample management is not just applied to compounds. It can be cell lines, it could be peptides, it could be even consumables.

And the more organized you are, the more likely you are to be functioning at a higher level as an organization.

But for sample managers, I encourage you to learn the machines, learn the softwares that drive the machines.

Don't just stick with one brand.

And the reality is with sample management it's such a high automation dependency job that the more you learn with automation the better tools you can apply to your solutions or your problems with unique solutions.

And so my biggest advice is go out there and learn different machines, learn different labware and learn different consumables and view the sales reps as your partners rather than your enemies who are trying to, you know take from you your organization.

Sales reps generally can be a help and the ones that are the best in this industry treat you as partners.

And so that's my recommendation is build relationships, learn more automation and you know, learn the technologies that you can apply to the unique problems in sample management.

Ruth Petersen

Perfect. Thank you so much for joining me today, Derrick.

Derrick Miyao

It was my pleasure.