Dr. Oznur Tastan will give a seminar on “Identifying Cross-Cancer Similar Patients via a Semi-Supervised Deep Clustering Approach” on 20 December at 13:40. The abstract of the talk and a short bio is shared below.

Bio:
Since 2018, Dr. Oznur Tastan has been with the Sabanci University Computer Science and Engineering, and Molecular Biology, Genetics, and Bioengineering departments. Before joining Sabanci, she worked as a faculty member at Bilkent University, Department of Computer Engineering, a post-doctoral researcher at Microsoft Research New England Lab (Cambridge, MA, USA). Dr. Öznur Taştan holds a BSc in Biological Sciences and Bioengineering from Sabanci University and received her MSc and her Ph.D. from Carnegie Mellon University, School of Computer Science, Language Technologies Institute. She has worked on diverse problems in computational biology; her present efforts center on building machine learning models to advance the current understanding of complex diseases. She is a recipient of the Young Scientist Research Award of the Science Academy (BAGEP), Tubitak Career Award, the UNESCO-L’OREAL National Fellowship for Young Women in Life Sciences, and METU Prof. Mustafa Parlar Foundation Research Incentive Award.

Abstract:
With the characterization of cancer tumors at the molecular level, there have been reports of patients being similar despite being diagnosed with different cancer types. Motivated from these observations, we aim at discovering cross-cancer patients, which we define as patients whose tumors are more similar to patient tumors diagnosed with another cancer type. We develop DeepCrossCancer to identify cross-cancer patients that always co-cluster with the other patient from another cancer type. The input to DeepCrossCancer is the transcriptomic profiles of the patient tumors, the age, and the sex of the patient. To solve the clustering problem, we use a semi-supervised deep learning-based clustering method in which the clustering task is supervised by cancer type labels and the survival times of the patients. Applying the method to patient data from nine different cancers, we discover 20 cross-cancer patients that consistently co-cluster. By analyzing the predictive genes of the cross-cancer patients and other genomic information available for the patient such as somatic mutations and copy number variations, we identify striking genomic similarities across these patients providing support. The detection of cross-cancer patients opens up possibilities for transferring clinical decisions across patients at a single patient level.