Professor Anna Lysyanskaya of Brown CS and Professor Ritambhara Singh of Brown CS and Brown University's Center for Computational Molecular Biology have each just received a Richard B. Salomon Faculty Research Award. The award, given annually by Brown’s Office of the Vice-President for Research, was established to support excellence in scholarly work by providing funding for selected faculty research projects of exceptional merit with preference given to junior faculty who are in the process of building their research portfolio.
“My research,” Anna explains, “focuses on one question: can contact tracing be automated, and with minimal consequences to privacy? Every hour, your phone creates a temporary pseudonym picked essentially at random that it broadcasts via Bluetooth to all devices within about a six-foot radius and receives pseudonyms from nearby phones. With an infected individual’s consent, public health authorities can take all pseudonyms their phone has emitted over a given period and publish them in a database without any identifying information. At that point, another smartphone can check whether it's been in contact with that phone, and if so, notify its user of the exposure without identifying the infected contact.”
However, in spite of the terrific benefits of this new technology, there’s also great room for improvement. “The privacy guarantees rely to a large extent on the promise that the device itself can’t be hacked," Anna says. "A hacked device will reveal more information to its user than just the date and duration of exposure; instead, it can reveal the precise time of exposure, enabling the user to figure out who the infected contact was. Additionally, a hacked device can wreak havoc by making false alerts possible. These privacy and security concerns have dampened the public’s buy-in into this approach.” In response, her project aims to create a set of cryptographic algorithms that eliminate both of these problems.
Ritambhara's research is centered on the availability of single-cell measurements, which provide a fine-grained heterogeneous cell landscape that reveals developmental trajectories across time for diverse cell types. "Studying these cell development trajectories gives us a better understanding of gene misregulation that leads to a diseased state in the cell," she says. "However, due to technical limitations, researchers can only observe this development at specific time-points or stages. We'll be using deep-learning models to fill this information gap by generating realistic in silico gene expression measurements. These measurements will be produced for missing time-points to augment the single-cell trajectory data, allowing improved downstream biological analyses. Recently, due to the generation of a large number of datasets, cutting-edge advancements in deep learning have been applied to the single-cell domain. However, the existing methods fail to factor in the temporal structure (time-point information) in the data – an important signal for observing cell development in single-cells."
Ritambhara's research group will model the temporal information in the single-cell gene expression. Their hypothesis is that the accurate modeling of the underlying biology will produce high-quality measurements for unobserved time-points. "Understanding how genes are regulated across space and time is an important question for researchers in the field (including at Brown)," Ritambhara says. "We aim to leverage the existing information using data-driven methods to help answer it for single-cell development."
For more information, click the link that follows to contact Brown CS Communication Outreach Specialist Jesse C. Polhemus.