FAQ

Negative Control is included in triplicate on each plate and consists of buffer run as a normal sample. These are used to monitor any background noise generated when DNA-tags come in close proximity without prior binding to the appropriate protein. The negative controls set the background levels for each protein assay and are used to calculate the limit of detection (LOD).

The Calibrator in Target 48 Cytokine (human) is a pooled human plasma sample with spiked in recombinant antigens for proteins with low endogenous levels to ensure that all 45 proteins are detected within Limit of Quantification.

The Calibrator used in Target 48 Mouse Cytokine consists of a buffer with recombinant proteins. In both products the Calibrator is used to adjust the predefined standard curve along the y-axis during development at Olink. The Calibrator is also included in triplicate on each sample plate and used to normalize all samples.

The Calibrator for Target 48 Immune Surveillance and Target 48 Neurodegeneration is plasma-free, and composed of a pool of recombinant proteins, equivalent to the biomarkers targeted by the panel.

LOD can be calculated using the LOD function in the Olink Analyze R package and added to your dataset. A detailed description of how LOD is estimated can be found in the Olink Analyze LOD tutorial.

Data is delivered in Parquet file format. Parquet is a free open source file format built to handle columnar storage data. The file cannot be opened in Microsoft® Excel. Olink recommends installing and using DBeaver to view, query and export Parquet data. If required Parquet files can be exported to CSV format. Data will be delivered via the Olink® Data Transfer application, an on-prem solution where data is transferred and stored within the boundaries of Olink’s environment. For any questions around Parquet file format or alternative methods of data delivery, please contact our support team.

Intensity normalization is typically recommended for well-randomized multi-plate projects, whereas Plate Control normalization is typically useful for single plate and non-randomized projects. It is always recommended to randomize samples in projects when possible.

Please contact our support team to discuss study design and ensure that the best normalization strategy is used for your project.

All data is normalized in the QC software and does not have to be manually calculated. NPX values in Explore 3072 and Explore HT are calculated from counts values. The counts values for each assay and sample is related to the counts for the extension control for that sample. This ratio is log2 transformed resulting in ExtNPX. The ExtNPX value is then either normalized by subtracting the median ExtNPX value of the Plate Controls for that assay (knows as Plate control normalization) or by subtracting the median ExtNPX value of all samples (excluding the control strip) for that assay (known as intensity normalization). Both normalizations are performed for each assay per plate.

All Olink Explore sequencing data needs to be processed using Olinks preprocessing software in order to translate the sequencing reads into counts. The software recognizes sequences pre-defined by Olink within each read and assigns each correct read to a sample and a protein (a count), and outputs a counts file containing the number of counts per protein for each sample. This output is used to calculate NPX values. The raw data/sequences themselves does not provide meaningful information without preprocessing.

Analysis service delivers NPX files, which contain both NPX and counts.

The CV of NPX is calculated per assay based on the NPX values of the Sample Controls and assumes log-normal distribution, using the log-normal distribution formula for CV. The CV of pg/mL uses the quantitative value of the sample controls to calculate CV as quantitative values in pg/mL are not log-normally distributed.

The Sample Control is pooled plasma samples. These are used as external control to estimate inter and intra precision.

The Sample Control is run in triplicates on Explore HT and duplicates on Explore 384/3072. For Explore HT the Sample Control is provided with the kit.

In general we consider NPX from the two platforms (qPCR and NGS) to be on different scales, so any direct comparison between the two should be done with care. We have seen in our internal qPCR versus NGS comparisons that the correlation is on average good, above 0.75, but some proteins have a low correlation. Proteins with lower correlations tended to have limited spread in at least one platform and/or were typically close to the limit of detection.

We would advise to plan the analysis in such a way that you do not rely on the direct comparison of NPX values between the two platforms. It should be sufficient in most cases to analyze data from each platform separately and then compare the results (p-values and directions). There must be compelling reasons for combining Target 96 data with Explore data, since the quality for some proteins is reduced, and thus the usefulness of direct comparisons will be protein dependent. If data must be combined, at least 24 overlapping bridging samples should be included and data combined using the same method as for combining Target 96 runs or Explore runs (median adjustment).

For any bridging normalization to be successful, the bridging samples need to represent the study samples well and need to be within the dynamic range of most assays.

Yes. It is recommended to randomize samples within and across plates. In the case of multiple samples per subject it is generally recommended to keep samples from the same subject on the same plate but randomize subjects across plates. For further information please see below.

Why do we recommend sample randomization?
There are several reasons motivating randomization of samples. Firstly, randomization of samples minimizes the risk of introducing bias that may confound downstream analysis to avoid that technical variation is mixed with biological variation. Secondly, Olink recommend intensity normalization for multiplate projects that is heavily dependent on the underlying assumption that samples are randomized across plates.

We do understand that total randomization of samples is not always feasible due to various reasons. In such cases intensity normalization must not be applied but alternative solutions are instead either plate control/inter plate control normalization or overlapping reference sample normalization (i.e. bridge normalization).

How to randomize your samples
The simplest way to randomize your samples and prepare a sample randomization scheme adopted to the Olink plate specific format is through using the R package Olink Analyze function olink_plate_randomizer(), for demonstration please see our tutorial on plate randomization. With this simple to use function you can easily select what Olink product you want to run and specify whether you have longitudinal samples that you be kept on the same plate, as well as more custom setups with altering number of samples per plate etc.

Once the randomization scheme has been prepared the distribution of samples across plates can easily be assesed visually by avariable variables e.g. collection site, sex, treatment etc. using the olink_displayPlateLayout() and olink_displayPlateDistributions() functions respectively.

Please contact our support team if you have any question about sample randomization.

The Plate Control consists of pooled plasma originating from healthy blood donors. These are used in data normalization to compensate potential variation between runs and plates.

The Plate Control is run in five replicates on Explore HT and in triplicates on Explore 384/3072.

The Negative Control for Olink Explore consists of buffer run as a normal sample. These are used to monitor any background noise generated when DNA-tags come in close proximity without prior binding to the appropriate protein.

The Negative Control is run in duplicates on Explore HT and triplicates on Explore 384/3072 and is used to calculate the limits of detection.

For Olink Explore 3072/384, results are obtained in 36 hours; hands-on time: <5 hours

There are 3 safe stopping points in the Olink Explore HT workflow:

1. After the PCR program, the PCR plates can be stored at -20°C for up to 2 weeks.
2. After pooling, the pooled PCR product can be stored at -20°C for up to 2 week
3. After Library purification, the purified libraries can be stored at -20°C for up to 4 weeks.

There are 3 safe stopping points in the Olink Explore 384/3072 protocol:

1. After the PCR1 program, the PCR1 plate can be stored at -20°C for up to 2 weeks.
2. After the PCR2 program, the PCR2 plate can be stored at -20°C for up to 2 weeks.
3. After Olink Library purification, the purified Libraries can be stored at -20°C for up to 4 weeks.

Olink uses NGS (Next Generation Sequencing) as a tool to measure the relative concentration of specific protein biomarkers in a sample. During the sample and library preparation, short stretches of DNA with a unique sequence for each protein biomarker are produced. After sequencing, the number of counts (reads) for a specific DNA sequence is proportional to the original protein biomarker concentration.

The panel covers fg/mL – ng/mL. Please refer to the validation data document for detection ranges of the single assays.

Yes, Olink Target 96 panels include four internal controls for quality control and data normalization. They enable technical control of each step in the PEA, as outlined below

• Incubation Control: Two non-human antigen (PE and GFP) with matching probes for data quality control and monitoring all steps of the run.
• Extension Control: IgG antibodies with matching oligo pairs for data normalization and monitoring the extension reaction, amplification and detection-step of the run.
• Detection Control: Synthetic double-stranded DNA for data quality control and evaluating final amplification and detection step of the run

With Olink’s Target 96  panels, results are analyzed by relative quantification, where protein expression changes are measured for each protein separately. Due to the nature of relative quantification, no internal standards are needed. The NPX is in a log2 scale, where a high value corresponds to a higher protein expression. Using NPX values, protein expression can be compared between groups of the same study.

N/A means “Not applicable”, i.e. a metric marked with N/A will never contain any actual number. For example, %CV Inter for Plate metrics will always be N/A since %CV Inter for a single plate is not defined. “no data” is displayed when there exists no data to calculate a metric. For example, %CV Inter for Panel metrics will display “no data” when the panel only contains one plate. You need at least 2 plates in order to calculate this metric.