Expert Insight: Proteomics on NGS: Game-changing immunoassay technology to increase multiplexing and throughput

Watch this on-demand webinar to learn how the Proximity Extension Assay technology from Olink could be a game changer for proteomics

27 Jul 2020

Dr. Ida Grundberg, Chief Scientific Officer, and Dr. Erika Assarsson, Head of R&D, Olink Proteomics

Protein biomarkers are increasingly used in clinical research to identify protein signatures that can predict outcomes, stratify patient populations, and provide insights into disease biology. Proteomics technologies have been restricted, however, by limitations in throughput, multiplexing level, specificity, and sensitivity.

Recent developments in multiplex technologies that enable high-throughput analysis of many proteins simultaneously, while using minimal volumes of biological samples, are accelerating the utility of proteomics.

Olink Proximity Extension Assay (PEA) technology featuring dual-recognition with matched DNA oligo-linked antibody pairs and qPCR readout has already been used to generate >160 million protein data points and ~400 peer-reviewed publications. Newly launched, Olink® Explore 1536 now takes this to an entirely new level, with PEA coupled to next-generation sequencing (NGS) readout enabling 1,472 proteins to be measured with just 3 µL plasma/serum and a throughput of 1.3 million protein data points/week on one instrument.

All this and more is discussed in our on-demand webinar with Dr. Ida Grundberg, Chief Scientific Officer, and Dr. Erika Assarsson, Head of R&D, Olink Proteomics, who explain how this new technology could be a game changer for proteomics.

Read on for highlights from the live Q&A session or register here to watch it any time that suits you.

Q: How do you see Olink Explore in the context of longer-established proteomics technologies like mass spec?

IG: In general, it’s a good complement to mass spectrometry for a broader discovery – together with our platform, we believe that you can get a more complete picture of the proteome. 
However, what we have heard from working closely with our mass spec collaborators are the challenges that they have with throughput of samples, the processing they require to be able to screen their samples, and the volume of samples needed – but that is the strong advantage with our technology. With Olink Explore going up to 3K and 4.5K proteins, I think our platform can perform well.

Q: The technology you described sounds almost too good to be true. Is there good external evidence out there to say that it really works?

IG: We believe so, but that's one of the reasons why we choose to be very transparent. So, we share all the internal data that we generate, but we have this great community that we have built where we share all the external data, all the publications that are out there, all the correlation and comparisons that we have done to other platforms. Over 173 million protein data points generated and well over 400 peer-reviewed publications so far provide good external evidence, and maybe it really is just that good!

Q: Have you tried to use less than 1 microliter in your assays? What are your limits in terms of sample size?

IG: Yes, we have for the Explore product. When we use the semi-automated system, we perform the first reaction in even lower volumes. So, we run the mosquito by SPT Labtech which takes 0.2 microliters of the sample into the reaction, which works well and going even below that, we have tested that with one acoustic instrument that did not work satisfactory, at least not for the feasibility that we performed. I think the limit will be when you reach too low a volume, you would simply have too few molecules for some proteins to get a very precise measurement.

Q: How big is the protein panel available? Can this be customized?

IG: Yes, the big panel includes roughly 1,500 proteins. We can customize the panel and that is shown here on this slide; we call that the focus panel. So, once you have discovered the signature through our discovery panels, we can design a focused panel only with those most significant candidates. So, yes, that's possible.

Q: How is the data which has been obtained from the melanoma checkpoint time study being published?

IG: It has not been published yet, it's still unpublished data. So, we are working on the first manuscript from the pilot study and we'll, of course, continue with this analysis as quickly as we can.

Q: Is it possible that both probes bind to the same non-specific protein and lead to false-positive signals? How do you determine that?"

IG: Of course, there's always a possibility, but I would say that the likelihood is very low because for each probe, you will have a very low risk of false findings and the same goes for the second probe. With that low likelihood, we have assessed this in many ways. So, we feel very confident with our specificity of the product, but yes, it's hard to say for sure.

Q: Have you tried using solid samples such as tissue?

IG: Yes, we have. When we lyse the tissue and take a microliter of that, it works well. We have done this in a number of studies, especially in the oncology space, and since we're using such small volumes, you don't need to have a big block of the tissue but just a tiny microbiopsy is sufficient.

Q: Do you prefer serum or plasma?

IG: We validate on both sample matrices and also on the validation data that we shared. You can see how they correlate with each other. So, in general, we have high detectability in both matrices. We have slightly higher detectability in serum. In some more extreme conditions, in severe cardiovascular diseases, for example, there are some markers in serum that can be highly elevated. In those cases, maybe more extreme cases, we recommend plasma, but both matrices will generate good data.

Q: How can the data generated with the qPCR and NGS platforms be compared?

IG: The output data from the qPCR and NGS will both reflect the relative differences from sample to sample within each assay. The way we handle this, the output data – the raw data coming from the NGS reader – we perform log transformation of the counts. After that, we perform very similar normalization procedures. With that, we also get very comparable data for relative quantification.

Q: Do you have a standard curve for each analyte?

IG: Yes, but for the absolute quantification panels, we use standard carves for each analyte, we run everything at the multiplex. For the data, we present where we assess the sensitivity of our relative quantification panels. We also generate the standard curve for each individual assay, but we run everything in multiplex during a validation. By doing so, we can establish all the sensitivity parameters such as LoD and the upper Limit of Quantitation from the dynamic range. So, it is the standard curve for each analyte, but it's always run in multiplex to make it feasible. Our validation data is on our website, you can go there and find the standard curve for each analyte. So the short answer is yes, but note that these validation curves are generated using recombinant proteins in vitro and can’t be used to calculate absolute concentrations in biological samples.

Watch this on-demand webinar here>>
SelectScience runs more than 10 webinars a month across various scientific topics, discover more of our upcoming webinars>>