Skip to Content »

Vancouver Bioinformatics User Group (VanBUG) » archive for November, 2016

 Jonathan Carlson

Talk Title:
“Project Premonition: Scaling up pathogen surveillance”

Date/Time:
Thursday, November 3, 2016 6:00pm

Affiliation:
Senior Researcher, Microsoft Research

Web-site: Jonathan Carlson

Abstract:
Emerging and re-emerging infectious diseases are a central public health concern, with implications for economies and even the stability of nations. The current system for detecting pathogens is entirely reactive, reliant on aggregating local reports by clinicians, and typically takes weeks of symptomatic human infections before detection. Project Premonition aims to detect pathogens before they cause outbreaks, using mosquitoes to sample pathogens in the environment. Turning a mosquito into a device requires new technologies to autonomously locate, robotically collect, and computationally analyze mosquitoes. In this talk, I will provide an overview of Project Premonition, and present some recent results from field trials using our autonomous traps. Most of the talk, however, will focus on what we do after we collect the mosquitos, which is the effort I lead. Once we have mosquitos, we deep sequence them, then attempt to identify the identity of the host, as well as any viruses, bacteria, or parasites detectable in the bloodmeal. This is essentially a pan-metagenomics problem, in which source DNA may have derived from any vertebrate, plant, virus, bacterium or fungus, and for which the primary host genome (the mosquito’s) is poorly characterized. We have therefore designed a high-throughput metagenomics pipeline that uses all of RefSeq (and many additional viral databases) as it’s reference. To deal with the sparsity of RefSeq relative to global species diversity, we must identify distant matches, but we must simultaneously distinguish among highly homologous sequences. To this end, we use a mixture model approach to share information across reads. Currently, the end-to-end pipeline runs in <10hrs, and has been applied to genetic samples from mosquitos, ticks, birds and other challenging scenarios. I will discuss the algorithmic approach, early results on synthetic and real data, and key challenges and opportunities

Bio:
I am a senior researcher at Microsoft Researcher, where I’ve been in various groups since 2008. I received my PhD in Computer Science from the University of Washington in 2009, and my undergraduate degree in Biology and Computer Science from Dartmouth in 2003. My research interests have also been focused on computational biology. Since my graduate days, I have largely focused on viruses and immunology, with a special focus on HIV. Specifically, most of my work begins with the observation that the high genetic variability of HIV provides the substrate for rapid adaptation, which in turn provides a window into the immune response. Following this paradigm, I have published over 70 papers in the field, covering topics such as HIV transmission, cellular and innate immune responses, and vaccine design. Project Premonition thus provides an interesting shift of focus for me, though I continue to pursue various HIV projects as well. More information can be found at http://research.microsoft.com/~carlson, and at the Project Premonition page: https://www.microsoft.com/en-us/research/project/project-premonition/

Please note:
Trainees are invited to meet with the VanBUG speaker for open discussion of both science and career paths. This takes place 5:00-5:45pm in either the Boardroom or Lunchroom on the ground floor of the BCCRC

Recommended Readings

——————————

Introductory Speaker:
Raunak Shrestha (PhD Candidate, Dr. Colin Collins’ lab, Vancouver Prostate Centre))

Title:
“HIT’nDRIVE: Patient-Specific Multi-Driver Gene Prioritization for Precision Oncology.”


——————————

Webcast Link:
http://vidyoreplay.computecanada.ca/replay/webcastShow.html?key=TAKdbjoDvIqKdLn
(This technology is brought to you by Compute Canada and WestGrid with support from PHSA Telehealth)