Usage

OncoMark provides a Python API and a web interface to quantify hallmark activity from transcriptomics data. This page outlines how to use both interfaces effectively.

❗ Do not use the raw model file from GitHub directly — the Python API or web-server is required to ensure preprocessing are done correctly.

---

Example Data

You can find example input data for testing in the test directory of the GitHub repository:

OncoMark Test Data

---

Python API

The Python API allows you to directly process data and predict hallmark scores using the pre-trained model provided with the package.

Step 0: Install Packages

pip install OncoMark

Step 1: Import Required Libraries

Before using OncoMark, ensure you have your transcriptomics data in a pandas DataFrame.

import pandas as pd
from OncoMark import predict_hallmark_scores

Step 2: Load Input Data

Load your transcriptomics data into a pandas DataFrame. The data should be formatted with samples as rows and genes as columns.

# Load input data as a pandas DataFrame
input_data = pd.read_csv('model_test_data.csv', index_col=0)

Step 3: Predict Hallmark Scores

Use the predict_hallmark_scores function to process the data and generate hallmark activity predictions.

# Predict hallmark scores
predictions = predict_hallmark_scores(input_data)

# Display the predictions
print(predictions)

Step 4: Save Results (Optional)

You can save the predictions as a CSV file for further analysis.

# Save predictions to a CSV file
predictions.to_csv('predictions.csv')

---

Web Interface

OncoMark also provides a user-friendly web interface for those who prefer not to write code.

You can use the hosted web server to upload your data and obtain hallmark activity predictions:

OncoMark Web Server

The comprehensive video detailing how to use OncoMark web-server is embedded below:

---

Notes and Tips

• Ensure that the genes is in the columns of the datasets.
• Genes missing from the input data but used during model training will be filled with zeros, while additional genes not used in training will be ignored.
• Check the Documentation for troubleshooting tips and advanced usage.

---

With this guide, you are ready to start exploring OncoMark's capabilities to quantify hallmark activity from your transcriptomics data.

---