README.md

Impepdom: Neural network builder for peptide ligand classification

---

Description to be added...

Installation

Step 1: System pre-requisites

If you're using macOS, install pipenv with Homebrew

$ brew install pipenv

More documentation: https://github.com/pypa/pipenv

Step 2: Clone repository and install requirements

$ git clone https://github.com/mikochen0107/mhc-1-immunopeptidome-characterization.git
$ cd mhc-1-immunopeptidome-characterization
$ pipenv install

Usage

Scenario 1: Train a model

Step 1: Create a script

Create a script in /scripts or a Jupyter notebook in /notebooks

Step 2: Example code

import sys
sys.path.append("..")  # add top folder to path
import impepdom

model = impepdom.models.MultilayerPerceptron(num_hidden_layers=2, hidden_layer_size=100)
dataset = impepdom.PeptideDataset(
    hla_allele='HLA-A01:01',
    padding='flurry',
    toy=True)

folder, baseline_metrics, _ = impepdom.run_experiment(
    model,
    dataset,
    train_fold_idx=[1, 2, 3],
    val_fold_idx=[0],
    learning_rate=2e-3,
    num_epochs=5,
    batch_size=128)

trained_model, train_history = impepdom.load_trained_model(model, folder)
impepdom.plot_train_history(train_history, baseline_metrics)

Step 3: Run code (if it is a script)

$ pipenv shell
$ cd scripts
$ python <NAME_OF_SCRIPT>

Scenario 2: Hyperparameter tuning

To be added...

Data formats

To be added...