Tumors exhibit an evolutionary capacity, akin to natural selection, allowing them to survive, evolve, and spread, often resisting treatments. Artificial intelligence (AI) and machine learning tools are being utilized to understand the rules governing tumor progression at genetic, epigenetic, metabolic, and microenvironmental levels. Matthew G. Jones, an assistant professor at MIT, is developing computational approaches to create predictive models for cancer, aiming to improve patient outcomes.
Tumor Evolution and Treatment Resistance
A common issue in cancer treatment is initial patient response followed by treatment failure. This resistance primarily stems from tumors' inherent ability to evolve. This evolution involves complex changes in genetic makeup, protein signaling, and cellular dynamics, structurally altering the tumor system.
The Jones lab's central hypothesis posits that tumors follow predictable patterns in space and time. Their research is dedicated to decoding the molecular processes behind these transformations using both computational and experimental technologies.
The lab specifically investigates extrachromosomal DNA (ecDNA) as a crucial mechanism of tumor evolution.
These ecDNAs are circularized DNA particles excised from chromosomes, existing separately within the nucleus.
The Role of Extrachromosomal DNA (ecDNA)
Initially considered rare, extrachromosomal DNA (ecDNA) amplifications were later discovered to be far more prevalent in cancer than previously thought. Next-generation sequencing technologies have revealed their presence in approximately 25 percent of aggressive cancers, including challenging cases like brain, lung, and ovarian cancers.
ecDNA amplifications are critical because they enable tumors to adapt more rapidly to stresses and therapies.
This accelerated adaptability can significantly speed up disease progression in unexpected ways, posing a major challenge to treatment.
AI and Machine Learning in Cancer Research
Machine learning and artificial intelligence (AI) are central to the study of ecDNA amplifications and the broader mechanisms of tumor evolution. The ultimate goal is to seamlessly translate these laboratory findings into tangible benefits for patients.
One key technology employed is single-cell lineage tracing. This allows researchers to study individual cell lineages, precisely pinpointing when aggressive mutations emerge within a tumor's historical development. This rich historical data provides crucial insights into dynamic processes, thereby facilitating the development of strategies designed to intercept tumor evolution at critical junctures.
The research specifically aims to enhance patient stratification for drug responses, predict and overcome drug resistance, and identify entirely new therapeutic targets.
MIT's Collaborative Environment
Matthew G. Jones joined MIT primarily due to its unique integration of engineering and biological sciences. He specifically highlights the Koch Institute's innovative structure, which actively promotes collaboration between engineers and basic scientists, fostering a truly interdisciplinary research environment.
Jones also places significant value on MIT's strong emphasis on education and training. He views academic research as a vital service dedicated to developing the next generation of scientists.
Reflecting this multidisciplinary ethos, his lab is designed for a hybrid approach, seamlessly combining both computational and experimental disciplines.