“知彼知己，百战不殆” (If you know the enemy and know yourself, you need not fear the result of a hundred battles) – Sun Tzu, 544 - 496 BC.
The earliest descriptions of cancer date back to around 3000 BC, yet it was only about a century ago that cancer was recognized as a genetic disease. Traditional treatments such as surgery, radiation, and chemotherapy are still prevalent. While effective for some, these methods often do not precisely target cancer cells. The key to improving treatment outcomes involves understanding the molecular events that drive cancer, which can lead to more effective and less toxic personalized therapies such as the HER2 antibody and antibody–drug conjugate (ADC) treatments for tumors with high HER2 expression. Despite the advancements, our overall understanding of cancer remains limited, leaving many types without known molecular mechanisms or effective targeted treatments. My research focuses on extracting molecular mechanisms from various models from cell lines to patient samples, designing and performing experiments to verify these findings, and understanding the mechanisms of action to develop clinically feasible therapeutic strategies.
Selected Projects
CDK4/6 Inhibitors in NF1-depleted ER+ Breast Cancers
In ER+/HER2- early-stage breast cancers, about 20% show resistance to endocrine therapy, associated with low neurofibromin (NF1-low) levels. Multi-omic analyses indicate that NF1-low tumors exhibit increased CDK4/6 activity, making them particularly responsive to combined endocrine therapy and CDK4/6 inhibitors. This suggests that NF1 status could serve as a key biomarker for assessing treatment efficacy in clinical trials. We have also established IHC and mass spectrometry approaches that define a threshold of NF1 protein loss, which can benefit patient selection alongside somatic genomic test.
PKMYT1 in CDK4/6 Inhibitor-Resistant ER+ Breast Cancer
Endocrine therapies combined with CDK4/6 inhibitors are commonly used to treat ER+ breast cancer but often face drug resistance. We discovered that constitutively high levels of PKMYT1 expression are linked to resistance to endocrine therapies and CDK4/6 inhibitors, identifying it as a biomarker for resistance. Given PKMYT1’s critical role in the cell cycle, we combined a CDK4/6 inhibitor, lunresertib, with nucleoside analogs like gemcitabine to selectively target cancer cells resistant to CDK4/6 inhibitors and lacking functional p53.
Preclinical Bioinformatics: Targeting the Undruggable
The first generation of approved targeted therapies for driver mutations, like BCR-ABL and HER2, marked significant advances. Despite this progress, however, a large class of mutations remains traditionally “undruggable”—there are no medications specifically designed to correct the mutant genes. A new generation of therapeutic strategy targets not the mutations themselves, but the vulnerabilities these mutations create, to kill the cancer cells. By analyzing multi-omic data from preclinical models and clinical databases, I identify specific targets against tumors with the “undruggable” mutations using statistical models and machine learning, and further evaluate these targets based on scientific validity, the competitive pharmaceutical landscape, and internal pipeline coordination.
As the century-old wisdom teaches us, understanding our ‘enemy’—cancer’s molecular biology and genetic dependencies—and ‘ourselves’—human physiology, immunology, and clinical practices—is crucial. As a translational cancer biologist with expertise in molecular biology and clinical translation, and as a bioinformatician skilled in data analysis and interpretation, I am excitedly contributing to this vital field.