.. _tutorials:

Modeling Tutorials
===================

Examples of applying SQUID on previously-published deep learning models can be found in the following Google Colab notebooks.

    1. `Colab – Additive (local) surrogate modeling with DeepSTARR <https://colab.research.google.com/drive/12HR8Vu_8ji3Ac1wli4wgqx1J0YB73JF_?usp=sharing>`_
    2. `Colab – Pairwise (local) surrogate modeling with ResidualBind-32 <https://colab.research.google.com/drive/1eKC78YE2l49mQFOlnA9Xr1Y9IO121Va5?usp=sharing>`_
    3. `Colab – Variant effect (local) prediction with DeepSTARR–Kipoi <https://colab.research.google.com/drive/1wtpT1FF5nu1etTDOaV3A7ByXhuLqK071?usp=sharing>`_

Several Python examples are also provided in the SQUID repository.

    1. `Python – Variant effect (local) prediction with DeepSTARR–Kipoi <https://github.com/evanseitz/squid-nn/blob/master/examples/example_variant_effect.py>`_
    2. `Python – Additive (global) surrogate modeling with BPNet–Kipoi <https://github.com/evanseitz/squid-nn/blob/master/examples/example_global_additive.py>`_
    3. `Python – Pairwise (global) surrogate modeling with BPNet–Kipoi <https://github.com/evanseitz/squid-nn/blob/master/examples/example_global_pairwise.py>`_


.. note::

    While dependencies are automatically handled in the Colab notebooks, Python scripts must be run with the correct environment manually installed.