|
|
|
# Organising data
|
|
|
|
You are probably used to organsing data by storing it in nested folders or directories. A heirachical system like this is a very natural to organise your data in terms of the explanatory variables in your experiment.
|
|
|
|
For example, an experiment can be organised
|
|
|
|
|
|
|
|
` organism > treatment > replicate`
|
|
|
|
|
|
|
|
` date > organism > treatment `
|
|
|
|
|
|
|
|
There are some downsides to this approach:
|
|
|
|
|
|
|
|
##### Sometimes we don't know the relative importance of variable:
|
|
|
|
The final organisation of the experiment may be different from what was planned.
|
|
|
|
We the data is analysed is makes sense to organise the explanatory variables by their relative effect size.
|
|
|
|
If it turns out that the within organism variation is larger than the treatment effect size the analysis will require the data is organised by replicate rather than treatment.
|
|
|
|
|
|
|
|
##### We may no not know which variables are important:
|
|
|
|
Although an experiment was planed to examine a set of explanatory variables, it turn out that one of the extraneous variables we thought shouldn't affect the results turns out be important.
|
|
|
|
|
|
|
|
##### Exploratory analysis and meta-analysis
|
|
|
|
You may want to explore the relationships in a dataset that was not collected with that purpose in mind. For example, it may a 'opportunistic' dataset that was not collected as part of a controlled experiment, or you may be performing a meta-analysis that involves multiple datasets that were collected for different purposes.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
# Organising in OMERO
|
|
|
|
|
|
|
|
OMERO takes a rather different approach to organising data from the nest directories approach that allows more flexibility in how data is organised.
|
|
|
|
There are two types of containers (folders) in OMERO which are structured as
|
|
|
|
|
|
|
|
`Projects > Datasets`
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
As the hierachy is very flat it doesn't take have a very long list of projects which hard to navigate. ***The key to organising data in OMERO is using tags and key-value pairs.***
|
|
|
|
|
|
|
|
## Tags
|
|
|
|
|
|
|
|
## [Key value pairs](./key-value_pairs) |
