We are excited to announce that the groundbreaking IBM MAMMAL (Molecular Aligned Multi-Modal Architecture and Language) model is now available for free through the public API at curecancerwithai.com. Developed by IBM Research, MAMMAL is a multi-modal biomedical foundation model designed to support drug discovery and cancer research by integrating protein, small molecule, and gene/omics data into a unified framework. This capability is crucial for advancing our understanding of cancer biology and identifying potential therapeutic targets.

What is MAMMAL?

MAMMAL is a state-of-the-art multi-task model that has been trained on an impressive dataset of roughly 2 billion biological samples. By aligning information across proteins, small molecules, and single-cell gene/omics data, it offers a comprehensive view of biological interactions and mechanisms. This multi-modal approach is significant because it allows researchers to tackle complex questions in drug discovery and cancer research, including how different biological entities interact and how they can be targeted for therapeutic purposes.

Research Behind MAMMAL

The capabilities of MAMMAL are detailed in the research article published in npj Drug Discovery (Nature), titled “MAMMAL — Molecular Aligned Multi-Modal Architecture and Language for biomedical discovery”. The model was evaluated on 11 diverse drug-discovery tasks, achieving new state-of-the-art results on 9 of them, with performance comparable to existing state-of-the-art methods on the remaining tasks. The extensive training on billions of samples equips MAMMAL with the ability to generate predictions that can significantly enhance the drug discovery process and contribute to our understanding of cancer mechanisms.

MAMMAL in the Cure Cancer With AI API

Cure Cancer With AI now exposes several powerful MAMMAL prediction capabilities through its free public API. This API includes three key endpoints that researchers and developers can use to gather insights into protein interactions, drug-target dynamics, and clinical toxicity predictions:

• Protein–Protein Interaction Prediction (PPI)
  POST /api/v1/mammal/ppi
  This endpoint requires two amino-acid sequences as inputs (labeled protein_a and protein_b). It returns a binding-affinity class label, where "1" indicates that the proteins interact and "0" indicates non-interaction. This capability is crucial for understanding biological pathways and potential drug targets.
• Drug–Target Interaction Prediction (DTI)
  POST /api/v1/mammal/dti
  For this endpoint, you will need to provide a target protein's amino-acid sequence (labeled target_seq) and a drug in SMILES notation (labeled drug_seq). It returns a predicted pKd (−log₁₀ Kd), where a higher value indicates a stronger predicted binding affinity between the drug and the target protein. This information is essential for drug discovery efforts aimed at identifying effective therapeutic candidates.
• ClinTox Clinical-Trial Toxicity Prediction
  POST /api/v1/mammal/clintox
  This endpoint takes a compound in SMILES notation (labeled smiles) and predicts its toxicity. It returns a prediction label (1 for toxic/likely to fail trials, and 0 for not toxic) along with a raw score. Understanding potential toxicity is critical in the clinical trial phase, helping to filter compounds that may pose risks to patients.

All of these MAMMAL endpoints are available for free use on curecancerwithai.com. You can create a free API key at /api-keys and access comprehensive documentation, including parameters and code samples, at /developers. The free tier allows for 100 requests per hour per key, making it accessible for researchers and developers looking to leverage AI in their work.

Start Using It for Free

We invite you to explore the potential of IBM MAMMAL through our free public API. To get started, simply create your free API key at /api-keys and refer to the detailed documentation at /developers. With this access, you can make informed predictions about protein interactions, drug-target affinities, and clinical toxicity, all at no cost.

In conclusion, MAMMAL's advanced inference capabilities are now free to use on curecancerwithai.com, providing researchers with valuable tools to further their drug discovery and cancer research endeavors. Please remember that while these predictions can generate insightful research signals, they are not medical advice, and users should not consider them a substitute for clinical judgment or decision-making.

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.