At Cure Cancer With AI, we are excited to announce the availability of IBM's MAMMAL (Molecular Aligned Multi-Modal Architecture and Language) biomedical foundation model through our free public API. With this cutting-edge tool, researchers now have access to advanced AI predictions that can significantly enhance drug discovery and cancer research efforts.

What is MAMMAL?

MAMMAL is a state-of-the-art multi-modal biomedical foundation model developed by IBM Research. It is designed to integrate and analyze various types of biological data, including proteins, small molecules, and single-cell gene/omics data. By utilizing a unified sequence framework, MAMMAL can efficiently represent and process these different modalities, making it a powerful tool for biomedical discovery.

The significance of a multi-modal model like MAMMAL lies in its ability to tackle complex biological questions that require insights from various data types. This is particularly relevant in cancer research and drug discovery, where understanding the interactions between proteins and small molecules can lead to breakthroughs in treatment strategies.

Research Behind MAMMAL

The capabilities of MAMMAL are detailed in the research paper published in npj Drug Discovery (Nature). The model was trained on approximately 2 billion biological samples and evaluated across 11 diverse drug-discovery tasks. Remarkably, it achieved new state-of-the-art results on 9 of these tasks, demonstrating its effectiveness in predicting biological interactions.

This vast training dataset and the model's multi-task capabilities allow MAMMAL to excel in various applications, making it an invaluable resource for researchers aiming to explore the complexities of drug interactions and potential therapeutic targets.

MAMMAL in the Cure Cancer With AI API

Now, let’s delve into how you can leverage MAMMAL's power through our API. We offer three distinct endpoints, each designed to provide valuable predictions for your research:

• Protein–Protein Interaction (PPI)

  Endpoint: /api/v1/mammal/ppi

  By inputting two amino-acid sequences (i.e., protein_a and protein_b), this endpoint predicts whether these proteins interact. The output is a binding-affinity class label: "1" indicates that the proteins are interacting, while "0" signifies non-interaction.

• Drug–Target Interaction (DTI)

  Endpoint: /api/v1/mammal/dti

  This endpoint requires a target protein amino-acid sequence (target_seq) and a drug represented in SMILES notation (drug_seq). It returns a predicted pKd value (−log₁₀ Kd), where a higher score indicates a stronger predicted binding affinity between the drug and the target.

• ClinTox Clinical-Trial Toxicity

  Endpoint: /api/v1/mammal/clintox

  For toxicity predictions, simply provide a compound in SMILES notation (smiles). The output will include a toxicity prediction (with "1" indicating the compound is toxic or likely to fail trials and "0" indicating it is not toxic) along with a raw score for further analysis.

All of these MAMMAL inference capabilities are available FREE to use on curecancerwithai.com. You can create your free API key at /api-keys and access full documentation, including parameters and code samples, at /developers. The free tier allows up to 100 requests per hour per key, making it accessible for researchers and developers alike.

Start Using It for Free

To get started, visit /api-keys to create your free API key. You can explore the capabilities of the MAMMAL model through our API and read the full documentation at /developers. With 100 free requests per hour, you can seamlessly integrate these powerful predictions into your research projects.

Conclusion

The introduction of MAMMAL through our public API marks a significant milestone in making advanced biomedical AI tools accessible to the research community. This capability is now freely available on curecancerwithai.com, providing invaluable support to those working in cancer research and drug discovery.

As a reminder, while the predictions generated by MAMMAL are valuable research signals, they are not intended as medical or clinical advice. Always consult with qualified professionals when interpreting these predictions in a clinical context.

To dive deeper into practical AI-for-cancer-research updates, explore our latest blog posts, learn more about our mission, and see how you can support ongoing work on our donations page.

Cure Cancer With AI is an educational research and information platform. It does not provide medical advice, diagnosis, or treatment recommendations; always discuss care decisions with a qualified healthcare professional.