With the ever-increasing potential of new technology and the exponential growth of the life sciences field, researchers are always running into new problems to solve. In this interview series, we get scientists’ opinions on the ‘Big Challenge’ in their field and the steps being taken to address it. From new and unique hurdles to fresh takes on common problems, we dive into the complexities of the research landscape.
In this interview, we chat to Mathew Chamberlain (Johnson & Johnson Innovation Medicine) about the ‘big challenge’ in single-cell multi-omics and its impact on drug discovery.
FLG: What is your background and role?
Mathew: I specialise in early drug discovery in the pharmaceutical industry. I’ve worked in single-cell omics for about six years now, mostly for drug discovery efforts, often across autoimmune diseases, and sometimes immune-mediated diseases like oncology.
FLG: What is the ‘big challenge’ in your field?
Mathew: I think that the biggest challenge now is related to how new single-cell technology is. The quality and richness of datasets will get much better with time, like bulk RNA-sequencing. Right now, there could be many new advancements made in different diseases, organisms and scientific fields, but the technology is still undergoing a phase of growth and adoption. That’s related to the newness of the technology, and also to the cost and feasibility in adopting single-cell software and methods.
FLG: Why should people care about this challenge?
Mathew: To be blunt: it could be the difference between life and death. We need to continue investing in research and discovery tools, particularly new technologies that provide new insights, because there’s a proven track record of how innovation has led to the discovery of new medicines and technologies. Right now, if you have precious limited patient samples, and you want to get the most information out of that as you can, you would run single-cell technologies. Widespread adoption of single-cell technologies is a real chance to pave the way for novel therapies; it just needs cost-cutting, improved feasibility, alignment on best practices and establishment of a computational workflow.
FLG: What is being done to tackle the issue, or what should be done to tackle the issue?
Mathew: I think that some of the cost reducing methods definitely help. For example, some of the multiplexing methods that various groups have come out with recently. Every time I think about these questions, I just think about what happened with bulk sequencing 15 years ago, and then assume it’s going to happen a bit faster this time, because we’ve essentially done it once before. In bulk sequencing, it took about 10 years for the field to coalesce on a pretty set standard of computational workflows that most people are reasonably happy with. In single-cell, we’re basically starting that process now. The field is starting to coalesce a bit, which, together with reduced costs, will help standardize everything from sample collection to results.
Some aspects of multi-omics are far behind this in single-cell. I’ve seen a lot of spatial omics talks this year, which remind me of single-cell five years ago. In spatial right now, you have a few samples, and you’re observing something and it looks interesting and enticing. But, if you were any researcher in the field and had a dataset of 100 spatial samples today, before and after treatment, there’s no computational workflow to even answer those questions right now. You’d have to design it yourself. That’s where I see the field going, a lot of multi-omics integration workflows, where you’re putting together different data types, and building out larger and larger datasets and atlases. I see the field going that way in the near future.
Then I see the field moving more into spatial and then increasingly as the sample sizes get larger, more tissue layers and more omics layers get added. At that point, the number of available workflows goes almost to zero pretty quick. Then you have to start the clock again when the fields coalesce on a standard workflow. You need a statistical analysis plan going into it, but usually the omics data typically aren’t an endpoint, they’re just an observation that goes together with your trial. That’s why I mentioned cost and throughput as two things I would work on now, because then at least you’ll get observational data. And we’re starting to get that from trials that we run, and other pharmaceuticals are doing that as well.
FLG: What is your advice to people breaking into the field?
Mathew: The advice I have is that, going into it you’ll be surprised at how much you didn’t know based on bulk sequencing methods alone. It’s not so much the technical stuff – I mean how little we know about the underlying biology. For example, when we were studying Alzheimer’s disease at Biogen seven years ago, and we were doing bulk RNA-Seq, and we had 500 samples in multiple conditions, the ‘party line’ was, ‘look, we know that we don’t have single-cell resolution, we’re looking at a bulk ensemble of cells here. But it’s not like we’re comparing humans to primates, things are going to be broadly similar, patient to patient.’ That was wildly wrong. We had no idea how many cell types occur only in diseased tissue, and not at all in healthy patients. That’s what we’ve learned from a decade of single cell sequencing – how little we knew before.