The Broad Institute of MIT and Harvard
Read The ReleaseOur mission
United efforts to drive progress against pediatric cancer
The Broad Institute, Dana-Farber Cancer Institute and St. Jude Children's Research Hospital have launched a large-scale, $60 million collaboration to advance understanding of the biological basis of pediatric cancers, identify new vulnerabilities of these diseases and accelerate therapies globally.
Media releases
About Us
The PedDep Accelerator is leading a multi-pronged effort against this problem focusing both on expanding and extending known successful approaches as well as investing in exploratory science with transformative potential. Central goals of the project include:
Despite decades of progress, cancer remains the number one cause of death by disease for children in the U.S., with some cancers still essentially incurable.
Dependency screening
Developing and deploying CRISPR-based genome editing techniques to identify hidden vulnerabilities (dependencies) in a spectrum of high-risk childhood brain, solid and hematological malignancies.
Omics profiling
Leveraging emerging technologies to characterize the genetic and epigenetic landscape of pediatric cancers.
Model derivation
Developing model systems where none currently exist for high-risk childhood cancers that have poor outcomes.
Combinatorial compound screening
Identifying effective combination therapies and mechanisms of drug resistance, and shortening the timeline for developing new therapies.
Data science
Developing computational approaches to mine and integrate data and developing innovative software tools for data sharing.
The PedDep Accelerator team
We are an actively meeting group of more than 100 investigators, including faculty, staff scientists and trainees from all three collaborating institutions.
Project leadership
![]()
Charles W. M. Roberts, MD, PhD
Director, St. Jude Comprehensive Cancer Center Executive Vice President, St. Jude Children's Research Hospital
Visit institution website![]()
Kimberly Stegmaier, MD
Vice Chair of Pediatric Oncology Research, Dana-Farber/Boston Children's Cancer and Blood Disorders Center Ted Williams Investigator, Dana-Farber Cancer Institute Professor of Pediatrics, Harvard Medical School Institute Member, Broad Institute
Visit institution website![]()
Francisca Vazquez, PhD
Director, Cancer Dependency Map Cancer Program, Broad Institute of MIT and Harvard
Visit institution website
Related scientific resources
The PedDep Accelerator pursues exploratory science while building on successful projects at the Broad, Dana-Farber and St. Jude.
A First-Generation Pediatric Cancer Dependency Map
Pediatric cancers are known to harbor fewer somatic mutations than adult cancers. However, this landmark study of genome-scale CRISPR screening revealed a similar complexity of genetic dependencies in pediatric as in adult cancer models. This first-generation Pediatric Cancer Dependency Map, PedDep v.1, included CRISPR screening results from 82 pediatric cancer cell lines representing 13 solid and brain tumor types. Many pediatric dependencies were distinct from those found in adults, underscoring the importance of dedicated therapeutic discovery efforts for pediatric cancers.
Explore PedDepDiscover more
Platforms that empower the cancer research community
The Cancer Dependency Map Project
The goal of the Broad's Cancer Dependency Map (DepMap) project is to create large-scale datasets and tools to systematically capture cancer vulnerabilities. The DepMap portal brings together comprehensive datasets to enable integrative analyses within tools to test hypotheses and to power target and biomarker discovery.
Learn more
St. Jude Cloud
St. Jude Cloud provides genomic data from more than 10,000 pediatric cancers; genomic, clinical and patient-reported data from pediatric cancer survivors; and unique analysis and visualization tools for the global research community.
Learn more
Your contribution
Interested in contributing cancer models?
The PedDep Accelerator is accepting contributions of pediatric cancer cell lines and other relevant models for dependency screening. If you have a model you would like to contribute for possible inclusion, please contact us: