(Photo courtesy of Supportive Guru)
By: Nathaniel Tok, Peak Associate

 

SFU computer science professor uses data mining to improve healthcare

SFU Computer Science professor Martin Ester is using data to help prevent adverse drug reactions (ADRs), such as death and disabilities, in patients undergoing cancer treatments.

Young cancer patients face a much greater risk of ADR, with 42% developing “disabling or life-threatening ADRs from cancer treatment.”

Ester’s research team, headed by Dr. Bruce Carleton and Dr. Colin Ross, is based in the B.C. Children’s Hospital, and is studying how genetics influence how patients might respond to drugs. They received a $9.9 million grant from Genome Canada for the research.

Within the team, Ester is using machine learning to data mine the clinical and genomic dataset, collecting DNA samples and recording reports of drug use and ADR outcomes.

According to Ester, “this [technique] will help us identify new connections between genetic variation and ADR” to help predict the probability of ADR during cancer treatment and allow for personalized treatments to prevent ADR.

The team is looking at ways to apply their findings in the medical world. The hope is that their efforts will allow for accessible pharmacogenomic testing to improve individual and population health.

 

How do researchers use social media?

The research group led by SFU professor Juan Pablo Alperin in the Master of Publishing program is investigating which social media platforms are the best for distributing academic research to local and international populations, and how language plays a part in information distribution.

In a paper published earlier this year, the group looked at the Zika virus outbreak event in 2016 as a case study to examine the communication of academic scientific research to various populations, and how the research was discussed locally, nationally and internationally.

Upon collecting Zika-related Tweets and Facebook posts in the first half of 2016, the group used an algorithm to determine that the majority of the posts were written in English. 90% of the Twitter posts and 76% of the Facebook posts were written in English, despite the Zika outbreak being located in Brazil. In contrast, only 7% of Facebook posts and 1% of Twitter posts relating to Zika were written in Portuguese, Brazil’s national language. The authors of the paper used this finding to conclude that “scholarly findings about the Zika virus are unlikely to be distributed directly to relevant populations through these popular online mediums.”

The group also found that the probability of non-English social media posts falls when more researchers speak English.

However, between platforms, the group found that Twitter was better in reaching international audiences while Facebook was better when talking to local populations. “Our results suggest that Facebook is a more effective channel than Twitter, if communication is desired to be in the native language of the affected country,” concluded the researchers in the paper.