Public repositories contain millions of samples and thousands of studies spanning a wide range of biological questions, populations, technologies, and experimental designs. The challenge is rarely the lack of data. Instead, the challenge is identifying studies that are relevant, trustworthy, and suitable for a specific objective.
Study discovery is the process of transforming a research objective into a set of candidate studies that can support downstream analysis. Before any data are downloaded, researchers must determine which studies are most appropriate for their intended purpose.
Effective data acquisition begins with selecting the right studies.
For example, a project focused on constructing a healthy reference microbiome cannot rely on a simple keyword search alone. Researchers must determine:
The quality of study discovery directly influences the quality of the resulting dataset.
Every study discovery process begins with a clearly defined objective.
flowchart LR
A[Research Objective] --> B[Search Strategy]
B --> C[Study Discovery]
C --> D[Study Evaluation]
D --> E[Candidate Studies]flowchart LR A[Research Objective] --> B[Search Strategy] B --> C[Study Discovery] C --> D[Study Evaluation] D --> E[Candidate Studies]
A clearly defined objective helps focus repository searches and reduces the risk of collecting unsuitable datasets.
Include:
Examples of exclusion criteria include:
Clearly defined inclusion and exclusion criteria improve reproducibility and ensure consistent study selection.
Several repositories can be used during study discovery.
| Repository | Typical Use |
|---|---|
| NCBI | Broad study discovery |
| GEO | Functional genomics studies |
| ENA | Sequence archive exploration |
| MGnify | Microbiome studies |
| GWAS Catalog | Variant association studies |
Although multiple repositories may be consulted, NCBI is often the starting point for identifying candidate BioProjects.
Study discovery is fundamentally a metadata-driven process.
Relevant metadata may include:
Metadata availability often determines whether a study can be included in a downstream analysis.
Before selecting a study, consider the following questions:
Studies that satisfy these criteria can be shortlisted for metadata acquisition.
Build a healthy reference gut microbiome dataset.
flowchart TD
A[Healthy Reference Goal] --> B[Search Public Repositories]
B --> C[Identify Candidate Studies]
C --> D[Review Metadata]
D --> E[Apply Inclusion and Exclusion Criteria]
E --> F[Shortlist Studies]flowchart TD A[Healthy Reference Goal] --> B[Search Public Repositories] B --> C[Identify Candidate Studies] C --> D[Review Metadata] D --> E[Apply Inclusion and Exclusion Criteria] E --> F[Shortlist Studies]
Example candidate studies used throughout this guide include:
These examples demonstrate how study discovery can identify datasets that satisfy the same biological objective while representing different sequencing designs.
The primary output of study discovery is a shortlist of candidate studies suitable for downstream acquisition.
Research Question
↓
Repository Search
↓
Study Evaluation
↓
Candidate BioProjectsThese BioProject accessions become the starting point for accession exploration and metadata acquisition.
Once candidate studies have been identified, the next challenge is understanding how public repositories organize biological data. In the next chapter, we examine accession systems and the relationships between BioProjects, BioSamples, Experiments, and Runs.