Friday Hacks #182, Oct 18: SRE and Voice Based Cancer Genomics
Posted on by Raynold Ng
Date/Time: Friday, Oct 18 at 7:00pm
Venue: Town Plaza Seminar Room 5, UTown, NUS
Site Reliability Engineering Practice in ViSenze
Talk Description:
We are an AI company focus on visual search and image recognition, we serve our AI technical technology globally and consistently provide high-performance service. This talk will share some practice we have in ViSenze that help us build a reliability system, as well as how we manage our online infrastructure and integrate it with our day to day engineering development lifecycle.
Speaker Profile
De Jun is an Engineering Manager of ViSenze, join ViSenze for more than 5 years and oversee few engineering teams in ViSenze.
Voice-based cancer genomics with Melvin
Talk Description:
Advances in high-throughput sequencing technologies have been a mixed blessing for genomics. Large-scale data production — while a driving force for biomedical discovery — frequently outpaces our interrogation capabilities. Without implementing proper tools, the disparity between information and bona fide biological knowledge will continue to grow. The life sciences must take advantage of emerging technologies, particularly those toting artificial intelligence, to exploit rather than submit to the data onslaught. Here we will discuss Melvin, a virtual genome analysis assistant built using Amazon’s Alexa technology. Leveraging domain-specific knowledge, Melvin uses an intent-mapping scheme that allows users perform conventional cancer genomics queries with simple voice commands. In addition to vocal responses, graphical results can be delivered to users via the Alexa Presentation Language. Melvin’s current knowledge-base is an amalgamation of data compiled from popular resources such as The Cancer Genome Atlas and the Gene Tissue Expression project. We have transposed the hierarchical nature of these databases into a deterministic finite state machine; this allows users to intuitively and systematically traverse analytical levels such as projects, phenotypes, and data types. As such, Melvin can support interactive, exploratory analyses as well as queries that cannot be reduced to a single, coherent utterance. Overall, this work modernizes how we interact with genomics data and helps streamline scientific inquiries.
Speaker Profile
Dr. Jason Pitt received his Bachelor’s degree in Biology with Honors from Gustavus Adolphus College. After a two year post-baccalaureate fellowship at the Genetics Branch of the National Cancer Institute (USA), he went on to earn his Ph.D. in Genetics, Genomics, and Systems Biology from the University of Chicago. There, under the guidance of Dr. Kevin White, he 1) developed computational workflows to efficiently process high-throughput sequencing data; 2) explored interactions between germline and somatic variation in cancer, and 3) investigated the genomics of racial disparities in breast cancer. In 2018, he started a laboratory at the Cancer Science Institute of Singapore (NUS) as a Special Fellow. In addition to conducting large-scale genome analyses, his laboratory focuses on developing innovative software for the identification, harmonization, and interrogation of genomic datasets.
Please RSVP at our Facebook event here if you are coming so we know how many people to cater for.
See you there!