Tutorials

This page provides an overview of tutorials on how to get started using immuneML and instructions on how to use immuneML for various different use cases.

All immuneML analyses are specified using a YAML specification file. To learn how to construct this file, see this tutorial:

• How to specify an analysis with YAML

Each analysis begins with selecting the dataset that will be used. In immuneML, the user can choose to import an existing dataset, or to generate a dataset made of random sequences (for example to test out some functionality without needing to use a specific dataset, or as a benchmarking dataset). The respective tutorials can be found here:

• How to import data into immuneML

• How to generate a random sequence, receptor or repertoire dataset

Using the specified dataset, immuneML can be used for various purposes: one can train and assess an ML model for immune repertoire or receptor-level classification, perform an exploratory analysis (to run preprocessings, encodings and reports without training a ML model), or simulate immune events by implanting sequence motifs in the dataset. See the tutorials below:

• How to train and assess a receptor or repertoire-level ML classifier

• How to apply previously trained ML models to a new dataset

• How to perform an exploratory data analysis

• How to simulate antigen or disease-associated signals in AIRR datasets

Tutorials:

• How to specify an analysis with YAML
• How to import data into immuneML
• How to generate a random sequence, receptor or repertoire dataset
• How to train and assess a receptor or repertoire-level ML classifier
• How to apply previously trained ML models to a new dataset
• How to perform an exploratory data analysis
• How to simulate antigen or disease-associated signals in AIRR datasets
• How to discover disease- or antigen specificity-associated motifs