Spectral Encoding, Molecular Screening and the Future of Connected Data featuring Ramesh Ramji

How could you simplify imaging using boba tea and Star Wars analogies? Listen today. Welcome to The Connect Current, the flow of innovation in San Diego. We bring you the founders, investors, and leaders building the future across life science, tech, climate, defense, AI, and more. This is where San Diego's next chapter takes shape. Welcome to the Connect Current, the flow of innovation in San Diego. I'm your host, James Zanewicz, and this episode is being recorded at BioLabs San Diego. We want to thank BioLabs for their support because they're not only supportive of Connect, but they're an important part of San Diego's life science startup infrastructure, providing early-stage biotech companies access to lab space, community, and the resources they need to test, develop, and grow. But I'm I'm most excited because we're at Biolabs today, and I'm joined by Ramesh Ramji, founder and CEO of Spectral Therapeutics, a Connect Springboard company. Spectral Therapeutics is San Diego-based, and they're a biotech company developing spectral hyperplexing technology designed to push the limits of molecular screening. The company is working on new ways to improve encoding depth, speed, and data fidelity, and that's so that researchers can generate richer, more connected data for deep learning and discovery. At the center of Spectral Therapeutics' work are technologies like SpectraGel and SPELLs, or Spectrally Encoded Combinatorial Bead Libraries. In simple terms, the company is building tools that could help researchers screen complex molecular libraries with greater precision, speed, and information density. This is exactly the kind of conversation we like to have on The Connect Current because it brings together deep tech, biotech, data, AI, founder execution, and the infrastructure behind better discovery. It's also a great example of why programs like the Connect Springboard matter. They help founders working on highly technical, potentially transformative ideas sharpen their story, strategy, and path to market at multiple touchpoints across their growth. Today's conversation is about what Spectral Therapeutics is building, why it matters, and how spectral encoding can change what researchers are able to see and measure, And then most importantly, what partners, investors, and strategic collaborators should understand about your next chapter. So Ramesh, if someone's listening in and they've never heard of you or Spectral Therapeutics, what is the company building? What problem are you solving? And why did this feel like the right company for you to create?

First of all, thank you for having me over here, James.

Oh, it's great to have you here.

Yeah.

And we're sitting here with Trey, who always hides in silence.

And Trey too. He's here too.

Yeah.

So in simplest words, right, Spectral is actually building an optical encoding technology where you're able to track every particle in an experiment over and over with these optical barcodes. To make it even more simpler, have you ever tried boba tea or bubble tea?

Oh, yeah, yeah. Yeah?

So imagine each boba is roughly a few millimeters wide. You shrink that 1,000 times, and you're well aware of Star Wars.

So you make it into tapioca?

Oh, yeah. So let's say you have lasers from Star Wars with Master Yoda, right, that he carries. Shrink those lasers and put those into the boba.

Right.

And now you have optically encoded bobas. Okay. So what we do at Spectral is build chemical matter on these beads, on these bobas, so that they are traceable optically at an enormous scale. You could shrink all of these bobas into a small tube where we bring in miniaturization into play, and you can screen them with flow cytometers, which is roughly around 10,000 bobas a second and makes it much, much faster than what current technologies offer in the market, right? So that's what Spectral does at a fundamental level, and I'm happy to dig in more into it.

I'd love to dig in a little more because a lot of people will hear molecular screening and may not understand why it matters. So for a general audience, just to understand kind of what role does screening play in drug discovery, diagnostics, or just innovation in biotech, and where do today's tools fall short in a way that you're trying to fix?

That's a great question. So You could view screening as a funnel behind every medicine, right? So there are roughly around 4,000+ medicines in the market. We've got more than 10,000+ diseases. And with these there, we potentially have an accessible chemical space, which is roughly around 10 to the 60th. If I were to compare these in simpler terms, you look up at the night sky, guess how many stars are there?

I mean, infinite, yeah.

Yeah, infinite, right? So if you were to put a number to it, there are roughly around 10 to the 24 stars out there in the night sky, whereas as humans, we can potentially have access to 10 to the 60th potential medicines, right, that we could build. And so far, humanity has built around a few trillion, which is 10 to the 12 of these medicines using DNA encoding as a format for building molecular libraries. And at the same time, using high-throughput screening, which is a gold standard for molecular screening. We've built around a few hundred million compounds for these screening campaigns. So the question that we ask at Spectral is, these technologies that are out there, be it high-throughput screening or DNA-encoded libraries, they've hit a ceiling today. So screening overall is the funnel behind every medicine that we have access to. Before a medicine hits the market or even When a patient receives a medicine, any given scientist who has worked in that program has screened thousands and thousands of different potential chemical compounds that could be good medicines for that given disease, right? So I would classify screening tools as 2 fundamental tools that are available out there in the market. One is HTS, high-throughput screening, and the other one is DNA-encoded libraries. High-throughput screening requires massive real estate in terms of thousands of plates that need to be screened for a few million compounds, whereas DEL, it shrinks everything into a small tube where you do actually encode every single compound using DNA as a barcode to it, right?

Uh-huh.

We are changing that because we've noticed that HTS overall is slow. Let's say it takes roughly around 3 to 4 weeks to screen a million compounds. What we are envisioning at Spectral is if we could run these on beads, which are optically barcoded, we could screen them under 5 minutes. Right? So that's the—

Oh, wow.

Oh yeah. So that is the massive, massive technology leap that we are aiming to build. The same thing goes with DNA encoding. Although you do have miniaturization in play for DNA-encoded libraries, you actually do not gather as much data that you screen for. And to put it in simpler terms, let's say I have around a few million members of compounds in a given library, which are DNA-encoded. My actual final hits that I get from this library is only going to be around a few tens of molecules. The rest of the data, which is more than 99% of the library, is all lost using DNA-encoded library screening technology. We changed that using spectral encoding, where we have miniaturization in play to begin with, but at the same time, we are able to get 100% data coverage.

I love this. It's almost— you've leveraged this idea of boba tea, using it to kind of crash through the ceiling that that is there for previous technologies. And you're doing that with that spectral hyperplexing, and your technology is there for SpectroGel and SPELS. And so if we're digging in a little more, how does this approach work exactly and allow you to see, measure, and connect in ways that were harder before? What exactly does the science do that is so different and special?

Yeah, that's a wonderful question. So if you talk about optically barcoding as a technology itself, The best tool out there in the market today is offered by Luminex, where they do make kind of bobas as well, but then they use fluorophores to barcode these beads, polymer beads. The problem with fluorescence is that it has a pretty wide emission bandwidth. It's like a broad stroke of paint. What we do at Spectral is we use these laser particles which have ultra-narrow emission bandwidths. Which allows us to scale to many, many orders of magnitude than what Luminex offers today in the market, right? If I were to put strong numbers to it, the current multiplexing technologies that Luminex carries and offers is roughly around 500-plex to the max of 1,000-plex assay. What we do at Spectral, and we have proved so far, is a few hundreds of million-plex. on beads already. Our theoretical limit is in the order of few trillions of unique barcodes on beads, which allows us to build all these molecular libraries, uniquely barcode them with optical barcodes on beads, and screen them in light speed. Wow.

It's really fascinating the way we're looking at this. And you've leveraged AI for this as well, connected data, deep learning. So why are encoding depth, speed, and data fidelity so important as that goal that we're all trying to reach of generating better data for AI, machine learning, and discovery is there? And I know the importance of data, like if you don't put good data in, you don't get good outcomes. But can you explain a little more about that and how that intersects with kind of what you're doing with Light? Because I, if I want to take it down to the simplest, I feel like what you just said was what exists is a wide fat marker and we're that thin custom tip pen. In a way. And how does that then translate into better data for these things? I guess is my simple question.

Yeah, that's a good question too.

Was the analogy good? Did it make sense? Because it popped that way in my head as you were talking, and I kind of felt like, oh, what's there is this wide emission band marker, and you've got this narrow, able to find things on that really fine point.

Just to add on to that, every bead or every BOBA that we carry does not have just one fine point, but multiple fine points. It's like a note.

Like the jumbo box of crayons we had as a kid, only with fine points.

Exactly.

I like it.

All right. So the play with AI over here is we envision to produce better predictive data compared to other tools, faster and better prediction. AI today, if I were to compare AI and give it a simpler, meaningful comparison, right? A soup is only as good as the ingredients that it carries. If you have bad ingredients in it, you wouldn't like the soup at all.

Stone soup, not so great.

Exactly.

Yeah.

So the same thing with AI models too. You have data that trains deep learning models, and if this data is bad, then your models, of course, have poor validation rates. It's as simple as that. Our technology allows us to gather as much data as possible from any given screen, very similar to what HTS can give us. Unlike DEL DNA-encoded libraries, where you screen for a few millions to a few billions of compounds at the same time, you only end up getting data from a fraction of it, be it 0.1 or less than 1% of it, right? Whereas the rest, 99% of data is just wasted. We want to change that. The power of DNA encoding as a technology is the fact that you're able to combinatorially produce compounds, synthesize them, and encode them with DNA tags. We envision to do the same, but with optical tags on beads, right? So we call it solid phase optical encoding and chemical library synthesis. So having been able to screen beads at a miniaturized level and screen compounds against a given protein target, we are able to generate massive datasets looking at protein and small molecule interaction or protein peptide interaction. or given any biomolecule interaction. All this data will allow us to build better deep learning models. And that is our goal at the end of the day. With that, we should be able to build a better community where these predictive models behave better. And that's our goal.

As you've been on your way to that goal, you've been a Connect Springboard company. So part of that journey is sharpening how any complex technical platform connects to real customer, partner, and investor needs. And so as you build your next chapter, I'm curious what kinds of partners investors, our strategic collaborators would be most helpful for you right now? And what do you want those people to understand about the opportunity ahead if they work with you?

All right, so in terms of partners, any partner who has hard screening problem, the good news is we've had an opportunity to interact with multiple big pharma so far. They really love what we are building as a team. We're not doing this for the first time. We've done this before at 1859, which was a Connect company as well.

We love repeat customers.

Yeah, yeah, yeah. And this time around, we are bringing in the magic of optical barcoding altogether. Any pharma customer who have a strategic goal to miniaturize and optically barcode their entire chemical libraries, we are happy to chat with them about this. In terms of investors, deep tech investors who have patience to see such transformational technology hit the market is what we are looking for to partner with.

Given that you're looking for deep tech investors, have you met Marshall Fritz on our team yet?

No.

We will get that connected after this because Marshall came out of family offices that invested in deep tech and has joined us as our director of tech and defense. So he is someone you should get to know.

Sounds good. Looking forward to it.

So see, sometimes, you know, useful connections, you learn about new things Connect has as part of doing the Connect Current. I have loved this conversation so far, but we always like to wrap up with our standard kind of Broad, it can mean anything to you. You can choose how you take the direction of it question, but it's to close on our Connect Current note and ask you, what drives your innovative flow?

That's a good question. I would say working in an extremely multidisciplinary field with a visionary transformational product that we are building. It's not just one product. It's going to cover across not only therapeutics, but also diagnostics as well. It's a foundational product that we are building. So that is one. The other part is extremely mission-driven scientific repeat founders. Working along with them, that's fun, right?

So getting to work inside BioLabs and inside San Diego.

Exactly, yeah.

That's amazing.

BioLabs is also an amazing place to be at.

Abu will be very happy with that answer. You might have to have him listen and be like, hey, what can you do for me for a free month's rent? I don't know.

Exactly, yeah.

Oh my gosh. Well, I want to thank you so much, Ramesh, for sitting down with us. It was a fun conversation. I learned a lot. I love how you break science down into simple analogies like boba tea. You let me have my marker thing, which maybe was right, maybe wasn't. But I think it's a really great reminder that tools really do drive innovation. And they're not always— well, they're doing really complex things. It's often easy to understand. How that translates into the real world when you start bringing references to Star Wars and boba tea and things like that. And you're generating the data that's going to enable discoveries, not just for what you're doing, but for entire platforms of technologies down the line. So your impact is going to expand on a multitude of capabilities. And the fact that Spectral Therapeutics sits at that intersection of spectral encoding, molecular screening, deep learning, and founder execution is exactly why we wanted to have you on, why we're proud to have you as a multi-springboard Springboard founder, and let's be honest, why San Diego continues to be a place where you can do things like this, where deep tech and life science can intersect, and where you can pair that complex science with the understanding of the problem it can solve. And for you, it's not just one problem, it's a problem that affects many problems. So thank you again.

Of course. Thank you so much, James.

Thanks for listening to The Connect Current. If you want to learn more about To connect and explore resources across San Diego's innovation ecosystem, follow us on LinkedIn or visit connect.org. We'll see you next time on The Connect Current, where San Diego's innovation story keeps moving forward.

This podcast was edited and produced by Resonate Recordings.