Back to Technologies Available for Partnering

Molecular Oncology Almanac: A Comprehensive Genomic Interpretation Tool

Therapeutics
Diagnostics

Project Overview

Precision oncology requires tools that can interpret complex genomic data to guide individualized cancer treatment. Current methods often focus on single gene alterations, leaving broader molecular features and interactions underutilized.
Scientists at Dana-Farber and the Broad Institute have developed the Molecular Oncology Almanac (MOAlmanac), an open-source clinical interpretation algorithm that integrates first-order and second-order genomic features for more comprehensive clinical decision-making.
MOAlmanac is used to identify therapeutic strategies for cancer patients, generate translational hypotheses, and support precision oncology trials.
MOAlmanac is freely accessible for research and nonprofit use.

The advent of genomic profiling has transformed cancer care, enabling clinicians to identify genetic alterations that drive tumor growth and guide targeted therapies. However, the sheer volume and complexity of genomic data present significant challenges. While current clinical interpretation methods focus on “first-order” genomic alterations—such as somatic variants and copy number changes—these approaches often overlook “second-order” molecular features, such as mutational signatures, tumor mutational burden (TMB), and microsatellite instability. These broader features can provide critical insights into therapeutic sensitivity, resistance, and disease prognosis.

Additionally, many patients have tumors that are “variant-negative,” meaning they lack clinically actionable genomic variants. For these patients, traditional genomic interpretation methods may fail to identify treatment options. To address these challenges, there is a pressing need for tools that integrate diverse molecular data, consider higher-order genomic interactions, and provide actionable insights for both clinical care and translational research.

Technology

The Molecular Oncology Almanac (MOAlmanac), developed by Eliezer Van Allen, MD, Chief, Division of Population Sciences and the Chandra Nohria Chair for AI in Cancer Research at Dana-Farber, and Brendan Reardon, a computational scientist in the Van Allen lab, is a groundbreaking tool designed to address the limitations of existing genomic interpretation methods. It combines a clinical interpretation algorithm with a curated knowledge base to provide integrative insights into patient-specific molecular profiles.

Currently, users can browse more than 70 cancer types, searching for genomic data from six evidence sources from inferential to FDA-approved, including nearly 180 therapies.

A Trio of Tools:

MOAlmanac is not a single entity. Instead, it currently includes a suite of tools, each addressing specific needs in precision oncology:

Foundational Clinical Interpretation Tool: The core MOAlmanac platform provides clinicians with actionable insights by integrating multimodal genomic data. It has been implemented in systems like Veterans Affairs and Count Me In projects, offering tailored genomic interpretations for diverse patient populations.
AI-Powered Chat Interface: A ChatGPT-inspired application built on MOAlmanac’s structured database allows clinicians to query genomic data in natural language. For example, users can input a patient’s cancer type and mutations to receive recommendations for therapies based on regulatory approvals and clinical evidence.
Machine Learning Model: A cutting-edge machine learning model trained on Dana-Farber patient data predicts therapeutic responses by analyzing genomic profiles and treatment outcomes. This model represents a new frontier in precision oncology, offering insights based on real-world patient data.

Key Features:

First-Order and Second-Order Integration: MOAlmanac evaluates both first-order genomic alterations (e.g., somatic variants, copy number changes, germline mutations) and second-order molecular features (e.g., mutational signatures, TMB, microsatellite instability). This dual-layer approach expands the scope of actionable insights, particularly for variant-negative tumors.
Profile-to-Cell Line Matchmaking: The tool includes a module that matches patient profiles to cancer cell lines, leveraging preclinical data to identify potential therapeutic strategies. This feature is particularly valuable for generating hypotheses in translational research.
Global Applicability: MOAlmanac is tailored for use in different healthcare systems, accounting for country-specific regulatory approvals and clinical guidelines. For example, the tool has been adapted for use in Ireland, with functionality that reflects local therapeutic options.
Open-Source Accessibility: MOAlmanac is freely available for research and nonprofit use, with licensing opportunities for commercial entities. Its structured database and algorithms are designed to facilitate global collaboration and innovation.
Local-first: MOAlmanac can be used within precision oncology workflows without requiring the use of external APIs. This makes it particularly suited to deployment within firewalled systems (e.g., Veteran Affairs).

MOAlmanac has been rigorously benchmarked against first-order interpretation methods in retrospective cohorts, demonstrating its ability to generate additional therapeutic hypotheses. In a prospective precision oncology trial cohort, the tool nominated a median of two therapies per patient and identified therapeutic strategies administered in 47% of cases. These results highlight MOAlmanac’s potential to improve clinical outcomes and advance precision cancer medicine.

Further details:

Primary Paper: “Molecular Oncology Almanac: Integrative Clinical Interpretation of Multimodal Genomic Data” (Nature Cancer, 2021) – This paper describes the methodology and benchmarking of MOAlmanac.

Additional Resources: You can learn more about MOAlmanac at its homepage: https://moalmanac.org/. It is available on GitHub, Docker Hub, and the Broad Institute’s Terra platform.

MOAlmanac represents a significant advancement in precision oncology, offering a comprehensive, integrative approach to genomic interpretation. By combining first-order and second-order molecular features, leveraging preclinical data, and tailoring functionality to global healthcare systems, MOAlmanac addresses critical gaps in cancer care and research. Its open-source accessibility makes it a valuable resource for clinicians, researchers, and commercial entities alike. With its potential to improve patient outcomes and drive innovation, MOAlmanac is poised to play a central role in the future of precision cancer medicine.

Team Members: Eliezer Van Allen, MD, Brendan Reardon, Sabrina Camp, Helena Jun, Maha Shady

Applications

MOAlmanac has a wide range of applications in precision oncology:

Clinical Decision Support: The tool helps clinicians interpret complex genomic data, identify actionable mutations, and recommend therapies based on regulatory approvals, clinical guidelines, and preclinical evidence.
Translational Research: MOAlmanac generates hypotheses for therapeutic sensitivity and resistance, enabling researchers to explore novel treatment strategies in preclinical and clinical trials.
Global Precision Medicine: By tailoring its functionality to different healthcare systems, MOAlmanac addresses disparities in access to precision oncology resources. Its adaptability makes it a valuable tool for international collaborations.
AI-Driven Innovations: MOAlmanac serves as a foundation for advanced AI applications, including machine learning models trained on patient data and chat-based interfaces for genomic interpretation.

Opportunities

MOAlmanac is freely accessible for research and nonprofit use, Dana-Farber’s Innovations Office is actively seeking partners to collaborate on the further development and dissemination of MOAlmanac. Interested parties are encouraged to contact the office for more information.

Get in touch