Chapter 12 Data Analysis
There are many types of analysis you can perform with Tercen. To see the some example cases, please explore the collections.
Go to https://tercen.com/explore/tab/collections.
Each project in this page, is an example project and in each project is a workflow containing the data analyis steps. The steps usually contain a computation called an operator. For a complete list of operators see the operator catalog.
Typical steps taken during an ‘omics’ analysis are the following:
- create or join a team for the analysis
- create the new project inside the team
- upload the data to the new project
- create a workflow to perform the following
- overview of raw data
- quality check (outlier samples, batch effects)
- normalization (if there is a systematic effect)
- transformation (usually a log function)
- exploration
- heatmaps. PCA, tSNE, fold changes, profiles
- clustering
- statistical testing
- t-test, anova
- biological interpretation
- pathway analysis, enrichment analysis
- network analysis
The parameters you adjust for are:
- kicking out samples (outliers)
- normalization or not (depends on the data)
- statistical test (t-test or anova)
Overview of raw data
Use a the following visuals to get an overview.
- Heatmap visuals
- Profile visuals
Quality check (outlier samples, batch effects)
There are a range of quality checks Here are some examples:
- Heatmap visuals
- Box plots visuals
Normalization (if there is a systematic effect)
There are a range of normalization techniques Here are some examples:
vsn_operator
normalizer_operator
Transformation (usually a log function)
There are a range of transformation techniques Here are some examples:
log_operator
asinh_operator
Exploration
There are a range of exploration techniques Here are some examples:
- heatmap visuals
- dimension reduction (PCA, tSNE, UMAP)
- fold changes
- profiles
- clustering