Login with your Reed credentials to view all events.

3203 Southeast Woodstock Boulevard, Portland, Oregon 97202-8199

http://academic.reed.edu/biology/seminars/index.html
View map Free Event

This Friday's Biology Department Seminar speaker is:

T.M. Murali, Ph.D.
Virginia Tech, Department of Computer Science,
Co-Director, Center for Systems Biology of Engineered Tissues

“XTalk: a Path-Based Approach for Identifying Crosstalk Between Signaling Pathways.”

Friday, March 24, 2017
3:50 pm Pre-seminar tea
4:10 pm Seminar begins
Biology 19

Student Lunch: Students interested in joining Dr. Murali for lunch should RSVP by emailing Kristy Gonyer ASAP since spaces are limited. Then plan to meet in B-115 shortly before noon on Friday. Commons dining room vouchers will be distributed at that time.

Abstract: Computational analysis of protein networks is pervasive in systems biology. However, the problem of using such networks to predict when signaling pathways may crosstalk has received very little attention. Existing computational methods that may be applied to discover such pathway pairs rely on simple overlap statistics between the proteins and interactions in the two pathways. We present XTalk, the first path-based approach for identifying pairs of pathways that may crosstalk. XTalk starts from the biological definition of crosstalk: when the stimulation of one pathway's receptors triggers a response downstream of the transcription factors of a different pathway. XTalk uses an efficient dynamic program to compute the exact statistical significance of the average length of multiple short paths that connect receptors in one pathway to the transcription factors in another. By design, XTalk reports the precise network of interactions and mechanisms that support the identified crosstalk. To evaluate XTalk, we manually curated the first ever gold standard dataset of 132 crosstalking pathway pairs and a set of 140 pairs that did not crosstalk. XTalk achieved an area under the ROC curve (AUC) of 0.65, a 12% improvement over the closest competing approach. The AUC varied with the pathway, suggesting that crosstalk should be evaluated on a pathway-by-pathway level. To conclude, we discuss the difficulties in creating an accurate gold standard and evaluating crosstalk algorithms and propose ideas for addressing these challenges.

Event Details

0 people are interested in this event