Olink Data Science Services

Study Design

Plan a well-designed, adequately powered protein biomarker study.

Data science training

Build confidence in analyzing and interpreting Olink proteomics data.

Reporting services

Receive expert reports that support biological understanding and decisions.

Data Science Tools

Access tools and resources across the complete data journey.

Study design

Why study design is important

The Olink Data Science Team are here to help you plan and design your study in the best possible way to help support a well-designed project and facilitate the generation of biologically meaningful findings.

The team are ready and willing to step in and help at any time during the study and data journey, but it is strongly recommended to involve them at the earliest possible stage of the process to ensure an efficient, optimally planned study.

TALK TO A SPECIALIST

A well-designed study sets a solid base to generate reliable, interpretable, and biologically meaningful data. This is particularly relevant for complex “omics” research where the volume and complexity of data could potentially lead to misleading conclusions if not properly controlled.

By starting with a clearly defined biological question and appropriate cohorts or other sample sources, the study can be designed to control biological and technical variability, minimize bias and enable appropriate statistical power. Scroll through the information below to see some of the major factors to consider in study design, from project objectives to study group selection, sample collection and handling, statistical power calculations, selection of protein panels and data analysis plan.

Study objective

Define what the proteomics data should help achieve.

Study Groups

Select suitable cases, controls, disease stages, or treatment groups.

Select appropriate cases & controls
Stratify by disease stage, treatment given, etc.

Sample size

Estimate the number of samples needed for statistical power.

Estimate expected effect sizes
Calculate samples required for adequate statistical power

Sample Handling

Minimize variation through planned collection, handling, and storage.

Select the optimal sample type
Minimize pre-analytical variation and document details
Randomize samples to avoid batch effects

Panel selection

Choose Olink panels aligned with the biology under investigation.

Focus on established and exploratory proteins
Ensure coverage of relevant pathways and functions
Prioritize biomarkers with low normal variation

Analysis plan

Plan normalization, statistical analysis, and biological interpretation early.

Perform optimal data normalization
Ensure robust conclusions through statistical analysis
Obtain biological insights through bioinformatics

Sessions:

Definition of Olink NPX data, reviewing approaches to normalization both within a project and across different projects.

Dimensionality reduction: Learning how to use PCA and UMAP functions both for data QC and clustering.

Introducing differential statistics.

Understanding pathway enrichment analysis and leveraging our online bioinformatics app, Olink Insight.

Predictive modeling approaches for exploratory biomarker research and evaluation of associations with study outcomes, and multiomics integration (this topic requires R programming skills).

Sessions:

Meta analysis with Olink data – learn how to apply meta-analysis to clarify how specific proteins are differentially expressed between groups across multiple studies.

Proteogenomic associations of protein quantitative trait loci (pQTL) and Mendelian Randomization (MR) – investigate genetic factors associated with protein changes and explore potential relationships with disease biology.

Network-based integration (WGCNA, PPI networks) – WGCNA and PPI analyses to provide a systems-level perspective by identifying networks of proteins that are co-expressed and working together, rather than examining them in isolation.

Data reporting

Reporting services

The dedicated specialists on the Olink Data Science Team provide comprehensive statistical analysis and data visualization services to propel your biomarker research forward. By partnering with our specialists, you receive quality data that supports research interpretation and analytical analysis. Two main types of data report can be provided, depending on your specific requirements: differential data analysis or normalization and bridging.

Differential data analysis

A report that gives an overview based on study objectives.

Differential expression analysis based on experimental groups such as different treatments, time points, age, sex or other variables.
Data quality control to determine sample outliers and evaluate global trends within the data.
Detailed protein expression plots stratified by group for significant assays, or predefined priority assays.
Volcano plots to visualize the difference in NPX-values between groups by significance for all assays.
Data visualizations that include a heatmap of significant assays, scatterplots evaluating sample variability, and a principal component analysis (PCA) plot to examine clustering of samples.

Volcano plot showing significant and non significant biomarkers comparing COVID 19 positive versus negative patients.

Download

Normalization and bridging

A report that provides you with adjusted NPX data normalized across studies run at different times.

An overview of the datasets, including a table of the overlapping samples and the Sample QC status in each project.
PCA plots and density plots that will be used to assess bridging quality by comparing plots pre- and post-bridging.
Violin plots for assays with high adjustment factors.
A newly compiled file with all datasets normalized towards the data from the original reference study.

A principal component analysis plot of adjusted NPX data normalized after bridging.

Download

Tools

Data science tools

From study planning to actionable insights, Olink provides biostatistical & bioinformatic expertise, tools, resources, and services to help accelerate your proteomics study and support you throughout your complete data journey.

Software by Olink

Explore tools that support study planning, visualization, and interpretation of complex proteomic data.

SOFTWARE OVERVIEW