Sold Out

Genome-scale signatures of gene interaction from compound screens predict c...

Event Information

Share this event

Date and Time

Event description


In-person session is located here:

Longwood Center 1053 | 360 Longwood Avenue, Boston, MA, United States

If you are located in Boston area, please come to our in-person session if convenient.

The time is 03/30/2018 Friday from 12:00 PM (EST) to 1:00 PM (EST).

The paper for this talk: Peng Jiang, Winston Lee, Xujuan Li, Carl Johnson, Jun S. Liu, Myles Brown, Jon Christopher Aster, and Xiaole Shirley Liu. In Press. “Genome-scale signatures of gene interaction from compound screens predict clinical efficacy of targeted cancer therapies.” Cell Systems.

The article is already online:

Time: 9:00 AM - 10:00 AM PDT (12:00 PM - 1:00 PM EST), Friday, March 30, 2018 link will be sent 24 hours before the event.

Please RSVP, the registration will be closed on 11:59 AM, March 28, 2018.


Identifying reliable drug response biomarkers is a significant challenge in cancer research. We present CARE, a computational method focused on targeted therapies, to infer transcriptomic signatures of drug efficacy from cell line compound screens. CARE outputs genome-wide scores to measure how the drug target gene interacts with other genes to affect drug efficacy in the compound screens. When evaluated using transcriptome data from clinical studies, CARE can predict the therapy outcome better than signatures from other methods. Moreover, the CARE signatures for the BRAF inhibitor are associated with an anti-PD1 clinical response, suggesting a common efficacy signature between targeted therapies and immunotherapies. When searching for genes in lapatinib resistance, CARE identified PRKD3 as the top candidate. PRKD3 inhibition, by both siRNA and compounds, significantly sensitized breast cancer cells to lapatinib. Thus, CARE should enable large-scale inference of response biomarkers and drug combinations for targeted therapies using compound screen data.

The source code and analysis results of CARE is available at

Peng Jiang

Dr. Peng Jiang is a postdoc research fellow at Shirley Liu Lab at Dana Farber Cancer Institute and Harvard School of Public Health. Peng finished his Ph.D. at the Lewis Sigler Genomics Institute at Princeton University, and he finished his undergraduate study with the highest honor at the department of computer science at Tsinghua University. Peng's research focused on developing computational models to identify biomarkers and regulators of anticancer drug resistance. Recently, Peng developed CARE, a computational method focused on targeted therapies, to infer transcriptomic signatures of patient clinical response from cellular compound screens (Jiang et al., Cell Systems 2018). The CARE model on targeted therapy efficacy can also be applied to determine the gene biomarkers of cancer immunotherapy response and resistance (Jiang et al., Nature Medicine in Revision). Peng also did many influential works on large-scale cancer data integration and biological network analysis. For example, the ENCODE consortium utilized his algorithm RABIT (Jiang et al., PNAS 2015) on the analysis of ENCODE3 genomics data and identified SUB1 as a new RNA binding protein in promoting tumor progression (Nature under review, Jiang as co-first author). Peng also developed a highly efficient network clustering algorithm SPICi, which has over 160 citations (Jiang et al., Bioinformatics 2010).

Strategic Alliance:




Date and Time

Save This Event

Event Saved