Noam Auslander, Ph.D.

In the past years, enormous growth in the quantity of biomedical data has been paralleled by advances in computational and deep learning techniques. These rapidly accumulating biomedical and clinical datasets have the potential to uncover unknown genetic drivers of cancer, carcinogenic infectious agents, and new treatment biomarkers. To achieve this potential, there is a need for improved and more interpretable computational strategies that can handle different types of biomedical data.

The Auslander laboratory develops advanced machine and deep learning methods to identify factors that drive cancer development and predict patients’ prognoses. We focus on the development of deep learning methods to identify new infectious agents in cancer, and the development of machine learning strategies that improve outcome prediction through biologically interpretable classifiers.

While the burden of cancer caused by infections is high, the attributable fraction is believed to be highly under-estimated by much of the cancer community. Identification of new cancer microbes has the potential to significantly reduce the global impact of cancer. However, identification of viruses or bacteria in tumors is highly challenging based short sequencing reads.

To address this challenge, our lab develops machine learning-based sequence analysis frameworks to better characterize the landscape of microbes that underlie different cancer types and correlate cancer-associated microbial species and proteins with prognoses and treatment responses.