Background image for our science section


Characterized human clinical samples

Belfer investigators are characterizing clinical samples to better understand correlations of tumor genetics and expression profiles of the tumor with the composition of the immune infiltrate. These extensive characterizations of the human disease help guide the selection and development of more accurate preclinical models for drug discovery and can identify relevant translational biomarkers.

Genetically engineered mouse (GEM) models of cancer

These complex mouse models reflect both the underlying genetic context and natural progression of the human disease. Because GEM models have a fully intact immune system, this platform is particularly well suited for in vivo evaluation of immuno-modulatory oncology drugs. Belfer has a large collection of relevant GEM models. Extensive characterization of these tumors and the immune cells that infiltrate them provide insights into the biology driving response and resistance to immunotherapies. 

"Liquid biopsies" for non-invasive disease monitoring

With advanced genomic analysis of patient plasma, Belfer Center researchers can identify and measure circulating tumor biomarkers and rapidly develop biomarker assays for clinical trials. These "liquid biopsies" help investigators assess drug response and rapidly detect and characterize drug resistance, all non-invasively.

Patient-derived xenograft (PDx) models

Generated from clinical samples collected during biopsy or surgery, these mouse models retain—with extremely high fidelity—the biological, molecular, and clinical features of their original tumor, including drug response profiles and genetic/genomic alterations. The Belfer Center's PDx models of non small-cell lung and high-grade serous ovarian cancers provide validated resources for biomarker and target identification, and evaluation of new single agent or combination approaches to therapy. 

Protein destablilization platform

Belfer Institute researchers are collaborating with Dana-Farber Cancer Institute investigators to refine and apply this cell-based approach, which utilizes changes in the stability and abundance of tagged proteins, including known oncoproteins, as readout for compound library screens. This platform can also reveal details of drugs' mechanisms of action and identify promising compounds active against traditionally "undruggable" targets, such as transcription factors.

Syngeneic models of cancer

These cell-line-based in vivo models complement human samples and GEM and PDx models, are easily scaled for larger preclinical studies, and are readily engineered to facilitate immuno-oncology experimentation. Belfer scientists are deeply characterizing existing models, as well as developing proprietary cell lines derived from our extensive library of GEM models.