Skip to content
Parker Institute for Cancer Immunotherapy
Search Close


A first-of-its-kind data analysis engine for scientific discovery.

Why This Research

Why haven’t we solved cancer with immunotherapy? A key reason is that we still don’t fully understand the complex interplay between cancer and the immune system. Without clarity, researchers cannot devise the right strategy against cancer.

But we’re getting closer. PICI scientists and many others are using advanced sequencing and imaging technology to analyze patient samples and get a holistic view of the action at the molecular level. In addition, there are thousands of clinical trials testing various combinations of immunotherapy in patients.

All these efforts generate hundreds of scientific papers each month – all available in the public domain. This makes public data a far greater resource than any single organization’s database.

So, what if there was a way to connect the dots across all the studies? A single system that can not only “see” the emerging and existing data, but perform advanced analysis so that critical scientific questions can be answered with up-to-date, accurate information?

We couldn’t find a platform like this. So we built one ourselves.

What We’re Doing

Our data scientists, immunologists, engineers and translational medicine experts designed the Cancer Data and Evidence Library (CANDEL) analysis platform.

The platform structures and aggregates data from published scientific papers combined with patient and assay data from PICI’s clinical trials and research projects. This single, integrated database allows for thoughtful cross-study analysis.

Powerful analysis tools are built on top of the platform. With data at their fingertips, data scientists can, with the push of a button, visualize the data and run machine learning models to draw insights in seconds.

Our platform has already begun to show its potential in clinical trials. With CANDEL, we are predicting who might respond to our novel immunotherapy combinations – a holy grail in the immunotherapy world. We’re still honing our precision, but the ability to detect early signals is significant. On top of that, we’ve achieved a fourfold decrease in the time spent structuring and analyzing data.

Ultimately, CANDEL turbocharges the work of our researchers to enable better, faster science.

The platform powers many of our projects at PICI. It guides our choices in what to study next and how we can advance big discoveries quickly. And, the platform will only become more robust as the field evolves.

With data from PICI and the field’s contributions, we imagine a world where each cancer patient can get the right therapy; where we can predict who will have major side effects; and where we can design therapies to help more patients.

CANDEL platform flowchart

About the Research

CANDEL empowers data scientists to do what they do best – efficiently gather insights from huge datasets.

Harnessing an innovative database technology called Datomic created by our partner Cognitect, the platform brings order to a vast amount of useful but messy data across sources both internal and external to PICI so that scientists can ask complex questions.

Raw data, whether it’s an imaging file or patient data, is loaded, structured, standardized, validated and stored in the database. Then data scientists can use CANDEL’s analysis tools to easily query, plot and model results to derive new insights.

The platform can currently process 17 types of molecular data, including genome and epigenome sequencing, gene expression, immune profiling of single cells in blood, serum proteins, imaging of tumor tissue and immune repertoire sequencing. CANDEL integrates this with clinical data including tumor measurements, patient survival, drug treatment regimens, demographics, clinical lab values, adverse events and more.

Data is also versioned, which makes experiments easily reproducible and ensures the integrity of our methods for years to come.

Where We’re at Now

CANDEL is the data-driven engine of our research. We started off by fueling the platform with publicly available data and data from our first human studies. Since then, the platform has grown to include nearly 40,000 samples from over 25,000 patients.

We currently aim to transform the way clinical research is done with this data at PICI. That means creating innovative treatment combinations, identifying which patients will benefit from which treatments and predicting adverse events and toxicities to manage them better.

In the future we hope to deploy CANDEL to the larger research community to accelerate the work of the entire field.