HICCC-logo blue.pngColumbia Cancer 50 Years Anniversary Logo
Big-data-wireframe-V2-dark.png
The Big Deal about Big Data
Transforming Research and Care with the Power of Data
Data can be found everywhere. It floods our news and social media feeds. It takes up storage in our smartphones that are in constant use. It drives innovation across computing, entertainment and finance. And in the world of cancer, the abundance of molecular data and patient health records generated and collected by the minute is already helping oncologists and scientists advance their knowledge of cancer and discover new therapies. And it’s just the beginning.
At the Herbert Irving Comprehensive Cancer Center, researchers are harnessing the information deluge to tackle critical problems in cancer and applying innovative computational approaches to advance research and transform care.
INTRODUCTION
Elham Azizi
ANDREA CALIFANO
Elham Azizi, PhD
Tracing the Path from Healthy Cell to Cancer Cell
Perhaps one of the most critical questions in cancer biology today is when does a healthy cell evolve into a cancer cell? That pursuit is what motivates biomedical engineer Elham Azizi, PhD, whose lab is devoted to uncovering the ‘whys’ and ‘whens’ of disease progression with the power of machine learning computational tools.
Classic machine learning algorithms tease out commonalties between objects or individuals, but Dr. Azizi designs new approaches better suited to study cancer biology: to identify the differences or outlier cases from genomic data.
Uncovering What's Unique about an Individual to Guide Personalized Therapies
“We're not focusing on trying to understand the average person or identify the most common patterns between data sets. We want to understand what’s unique about an individual to guide personalized therapies,” says Dr. Azizi, assistant professor of biomedical engineering and Herbert and Florence Irving Assistant Professor of Cancer Data Research (in the Herbert and Florence Irving Institute for Cancer Dynamics and in the Herbert Irving Comprehensive Cancer Center). “We want to know what is different in a patient’s biology that leads to disease.”
Dr. Azizi and her collaborators have developed DECIPHER, a novel computational framework that visually reconstructs the trajectory of a healthy cell to a cancer cell from the earliest stage. Focusing on underlying disease progression in acute myeloid leukemia (AML), DECIPHER analyzes single-cell sequencing data of patient tissue samples to model, step by step, how and when cancer cells derail from healthy development. The method uncovers developmental stages, or the order of events, that lead to transformation of stem cells from healthy states to cancer stem cell states, and then finally, to mature cancer cells. DECIPHER’s goal is to capture, computationally and visually, at which exact point a healthy bone marrow stem cell in AML patients takes a turn.
The Azizi lab is tackling a similar problem in melanoma with a method developed in collaboration with physician-scientist Ben Izar, MD, a member of the cancer center and assistant professor of medicine. ECHIDNA, a machine learning algorithm, tracks how diverse cancer cells evolve in response to therapies. By integrating genomic and single cell-resolution transcriptomic data from melanoma patients collected before and after treatment with checkpoint immunotherapy, they aim to characterize the subsets of tumor cells that persist with therapy and pinpoint resistance pathways they leverage to escape immune response.
“DECIPHER and the methods we are developing can guide the design of biomarkers for early detection of disease and improved strategies for immunotherapy,” says Dr. Azizi, a member of the Precision Oncology and Systems Biology Program at the Herbert Irving Comprehensive Cancer Center. “This problem is complicated and challenging and this is an exciting opportunity in cancer research where we can really make a difference.”
“We want to know what is different in a patient’s biology that leads to disease.”
—Elham Azizi, PhD
Assistant Professor of Biomedical Engineering and Herbert and Florence Irving Assistant Professor of Cancer Data Research
Elham Azizi, PhD
Tracing the Path from Healthy Cell to Cancer Cell
Andrea Califano, Dr.
A Predictive, Powerful Approach to Decipher Cancer
Andrea Califano, Dr.
A Predictive, Powerful Approach to Decipher Cancer
Humans have about 20,000 genes working together in ways that are different from cell to cell and from individual to individual. The enormous amount of data that has been collected on cancer has helped systems biologists like Andrea Califano, Dr., build computational models that, rather than trying to explain things in cancer one gene at a time, explain how all these genes work together in a system.
“There has been a dramatic revolution in the way we approach biological problems: It has gone from being an empirical-based process to being a data-driven and model-based discovery.”
—Andrea Califano, Dr.
Chair and Professor of Systems Biology, Vagelos College of Physicians and Surgeons; Co-leader of the Precision Oncology and Systems Biology Program, Herbert Irving Comprehensive Cancer Center
Dr. Califano, founding chair of the Department of Systems Biology at Columbia University Vagelos College of Physicians and Surgeons and co-leader of the Precision Oncology and Systems Biology research program at the Herbert Irving Comprehensive Cancer Center, has created a suite of computational tools that capture and analyze RNA-sequencing data of patient samples to predict and prioritize single drug or combination drugs that will work best for an individual patient and their specific cancer.
“Our systems biology approach focuses on the RNA [molecule] rather than DNA. Because RNA shows us specifically what is going on in a cancer cell, at a specific point in time, we have developed algorithms that accurately predict the proteins that drive the cancer cell, based only on RNA measurements,” says Dr. Califano.
“While this is more complex than looking for DNA mutations to target with treatment, it also promises to be more effective when it comes to dismantling cancer, because the protein activity state of a cancer cell provides the most informative data in terms of predicting whether a drug will kill it or not.”
The Califano lab has assembled sophisticated computational networks of molecular interactions between proteins and genes then analyze them to identify and target a handful of “master regulator” proteins, essentially the “pillars” that determine the cancer cell behavior and represent its most critical vulnerabilities. Their research has shown that these master regulator proteins work together to power the cancer cell, akin to a building standing up on a small number of load-bearing pillars; target one or more of these pillars and the entire building collapses.
“Our methodologies identify precisely which proteins in each cell are the load-bearing pillars of the cancer cell state and which drugs can best target their activities,” he says. The idea and goal, in the next frontier of precision cancer medicine, is to create even more powerful and predictive tools to determine the precise drug or drug combination per individual, per tumor, per cancer.
In essence, “Removing the guess work in treatment approaches that are truly personalized to the patient.”
big data 01 graphic
COLUMBIA CANCER ONWARD
Introduction
Nicholas Tatonetti
Elham Azizi
Andrea Califano
Big-data-wireframe-V2-dark.png
The Big Deal about Big Data
Transforming Research and Care with the Power of Data
Data can be found everywhere. It floods our news and social media feeds. It takes up storage in our smartphones that are in constant use. It drives innovation across computing, entertainment and finance. And in the world of cancer, the abundance of molecular data and patient health records generated and collected by the minute is already helping oncologists and scientists advance their knowledge of cancer and discover new therapies. And it’s just the beginning.
At the Herbert Irving Comprehensive Cancer Center, researchers are harnessing the information deluge to tackle critical problems in cancer and applying innovative computational approaches to advance research and transform care.
Data for Inclusivity
Over the last decade, bioinformatics has made a remarkable impact on cancer medicine and it has resulted in better outcomes, new treatments and longer survival. But these advances have been unequal in their distribution, says Nicholas Tatonetti, PhD, primarily benefitting white and wealthy patients.
Named Chief of Cancer Data Science at Columbia Cancer in May of 2022, Dr. Tatonetti is committed to using data science methodologies to improve an area of cancer care ripe for change: diversity in clinical trials.
“We don’t want to just operate successful cancer clinical trials, we really want to optimize our ability to enroll a diverse population in clinical trials,” says Dr. Tatonetti, associate professor of biomedical informatics at Columbia University Vagelos College of Physicians and Surgeons.
Along with high quality digital health records from cancer patients, genetic data is now adding to the repository of information, making it even more comprehensive. This represents a new opportunity to think about and evaluate how cancer clinical trials are being conducted, including how trials are being evaluated and how to reach out to and recruit patients into those trials so that scientists can accurately assess the effectiveness and safety in multiple groups and different groups that have traditionally been overlooked or marginalized.



Introduction
Nicholas Tatonetti
Elham Azizi
Andrea Califano
Elham Azizi, PhD
Tracing the Path from Healthy Cell to Cancer Cell
Perhaps one of the most critical questions in cancer biology today is when does a healthy cell evolve into a cancer cell? That pursuit is what motivates biomedical engineer Elham Azizi, PhD, whose lab is devoted to uncovering the ‘whys’ and ‘whens’ of disease progression with the power of machine learning computational tools.
Classic machine learning algorithms tease out commonalties between objects or individuals, but Dr. Azizi designs new approaches better suited to study cancer biology: to identify the differences or outlier cases from genomic data.
“With the unprecedented amount of data and tools available, we are able to identify the genes and mechanisms that drive each patient’s cancer.”
—Nicholas Tatonetti, PhD
Chief of Cancer Data Science at Columbia Cancer
Data for Inclusivity
Over the last decade, bioinformatics has made a remarkable impact on cancer medicine and it has resulted in better outcomes, new treatments and longer survival. But these advances have been unequal in their distribution, says Nicholas Tatonetti, PhD, primarily benefitting white and wealthy patients.
Named Chief of Cancer Data Science at Columbia Cancer in May of 2022, Dr. Tatonetti is committed to using data science methodologies to improve an area of cancer care ripe for change: diversity in clinical trials.
“We don’t want to just operate successful cancer clinical trials, we really want to optimize our ability to enroll a diverse population in clinical trials,” says Dr. Tatonetti, associate professor of biomedical informatics at Columbia University Vagelos College of Physicians and Surgeons.
Along with high quality digital health records from cancer patients, genetic data is now adding to the repository of information, making it even more comprehensive. This represents a new opportunity to think about and evaluate how cancer clinical trials are being conducted, including how trials are being evaluated and how to reach out to and recruit patients into those trials so that scientists can accurately assess the effectiveness and safety in multiple groups and different groups that have traditionally been overlooked or marginalized.



Uncovering What's Unique about an Individual to Guide Personalized Therapies
“We want to know what is different in a patient’s biology that leads to disease.”
—Elham Azizi, PhD
Assistant Professor of Biomedical Engineering and Herbert and Florence Irving Assistant Professor of Cancer Data Research
“We're not focusing on trying to understand the average person or identify the most common patterns between data sets. We want to understand what’s unique about an individual to guide personalized therapies,” says Dr. Azizi, assistant professor of biomedical engineering and Herbert and Florence Irving Assistant Professor of Cancer Data Research (in the Herbert and Florence Irving Institute for Cancer Dynamics and in the Herbert Irving Comprehensive Cancer Center). “We want to know what is different in a patient’s biology that leads to disease.”
Dr. Azizi and her collaborators have developed DECIPHER, a novel computational framework that visually reconstructs the trajectory of a healthy cell to a cancer cell from the earliest stage. Focusing on underlying disease progression in acute myeloid leukemia (AML), DECIPHER analyzes single-cell sequencing data of patient tissue samples to model, step by step, how and when cancer cells derail from healthy development. The method uncovers developmental stages, or the order of events, that lead to transformation of stem cells from healthy states to cancer stem cell states, and then finally, to mature cancer cells. DECIPHER’s goal is to capture, computationally and visually, at which exact point a healthy bone marrow stem cell in AML patients takes a turn.
The Azizi lab is tackling a similar problem in melanoma with a method developed in collaboration with physician-scientist Ben Izar, MD, a member of the cancer center and assistant professor of medicine. ECHIDNA, a machine learning algorithm, tracks how diverse cancer cells evolve in response to therapies. By integrating genomic and single cell-resolution transcriptomic data from melanoma patients collected before and after treatment with checkpoint immunotherapy, they aim to characterize the subsets of tumor cells that persist with therapy and pinpoint resistance pathways they leverage to escape immune response.
“DECIPHER and the methods we are developing can guide the design of biomarkers for early detection of disease and improved strategies for immunotherapy,” says Dr. Azizi, a member of the Precision Oncology and Systems Biology Program at the Herbert Irving Comprehensive Cancer Center. “This problem is complicated and challenging and this is an exciting opportunity in cancer research where we can really make a difference.”
“There has been a dramatic revolution in the way we approach biological problems: It has gone from being an empirical-based process to being a data-driven and model-based discovery.”
—Andrea Califano, Dr.
Chair and Professor of Systems Biology, Vagelos College of Physicians and Surgeons; Co-leader of the Precision Oncology and Systems Biology Program, Herbert Irving Comprehensive Cancer Center
Dr. Califano, founding chair of the Department of Systems Biology at Columbia University Vagelos College of Physicians and Surgeons and co-leader of the Precision Oncology and Systems Biology research program at the Herbert Irving Comprehensive Cancer Center, has created a suite of computational tools that capture and analyze RNA-sequencing data of patient samples to predict and prioritize single drug or combination drugs that will work best for an individual patient and their specific cancer.
“Our systems biology approach focuses on the RNA [molecule] rather than DNA. Because RNA shows us specifically what is going on in a cancer cell, at a specific point in time, we have developed algorithms that accurately predict the proteins that drive the cancer cell, based only on RNA measurements,” says Dr. Califano.
“While this is more complex than looking for DNA mutations to target with treatment, it also promises to be more effective when it comes to dismantling cancer, because the protein activity state of a cancer cell provides the most informative data in terms of predicting whether a drug will kill it or not.”
The Califano lab has assembled sophisticated computational networks of molecular interactions between proteins and genes then analyze them to identify and target a handful of “master regulator” proteins, essentially the “pillars” that determine the cancer cell behavior and represent its most critical vulnerabilities. Their research has shown that these master regulator proteins work together to power the cancer cell, akin to a building standing up on a small number of load-bearing pillars; target one or more of these pillars and the entire building collapses.
“Our methodologies identify precisely which proteins in each cell are the load-bearing pillars of the cancer cell state and which drugs can best target their activities,” he says. The idea and goal, in the next frontier of precision cancer medicine, is to create even more powerful and predictive tools to determine the precise drug or drug combination per individual, per tumor, per cancer.
In essence, “Removing the guess work in treatment approaches that are truly personalized to the patient.”