MSD Database Structure

MSD database has a hierarchical structure. We have the concepts of Projects, Organisms, Samples, and Datasets. As a user has defined a project, the new project can host various organisms, samples and datasets. A defined organism can have several samples assigned to and similarly each sample might have various datasets.

Take this hypothetical situation into consideration. As a researcher you might have project in which you have three mice. In this case each mouse would be an organism at MSD. Suppose that you plan to compare two types of samples (colon biopsy and feces) and two microbiome analysis methods (16S amplicon and Shotgun sequencing). In order to achieve this you need to attempt one sampling for each sample type and get enough sample for each sample type to send for different sequencing methods. It means that you would take 2 samples from each mouse and produce 2 datasets from each sample. Thus, you need to define all your mice as organism at MSD and for each of them define only 2 samples for which 2 datasets would get generated and uploaded to MSD. The table below shows the case.

Relation of entities at MSD

No.	Organism	Sample	Dataset Type
1	Mouse_1	Biopsy	16S Amplicon
2	Mouse_1	Biopsy	Shotgun
3	Mouse_1	feces	16S Amplicon
4	Mouse_1	feces	Shotgun
5	Mouse_2	Biopsy	16S Amplicon
6	Mouse_2	Biopsy	Shotgun
7	Mouse_2	feces	16S Amplicon
8	Mouse_2	feces	Shotgun
9	Mouse_3	Biopsy	16S Amplicon
10	Mouse_3	Biopsy	Shotgun
11	Mouse_3	feces	16S Amplicon
12	Mouse_3	feces	Shotgun

The table above shows that each mouse has 4 samples and each sample has 2 datasets. In total, there are 12 datasets

Note

Definition of a sample refers to each attempt of sampling. For example, two samples, both taken from feces, at the same time point would be considered two distinguished samples so that two different sample objects at MSD.