Science of Stories - Social Media Informatics on Genomic Technologies

Background

Social media is a place for information and ideas to disseminate. Some of these data can now be accessed through free and publicly available APIs. Twitter’s standard API, for example, provides access to 1% of total data with multiple extractable features about the user and the tweets including images, demographics and tweet contents. Similar data from Reddit retrieved through its API is also publicly accessible.

With careful management and reporting of such data to avoid violation of respective terms and conditions, many exploratory analyses can be performed to address questions related to health informatics and public perception of new medical technologies. Studies can also be scaled up using the paid Twitter premium API to access remaining data. Subjecting to the questions asked, analyses can be performed with a range of methods from basic statistics to deep learning.

It would be interesting to explore the public perception of genomic technologies.

Progress Report

Over 70k tweets related to genetic testing and related companies streamed during Black Friday weekend and Cyber Monday. Further analysis will follow.

A list of author keywords from over 2k publications related to genetic testing is retrieved using PubMed API. Further (network) analysis will follow.

2018 Global Strategic Business Report - £4,414 - A comprehensive report published in February 2018 on the key players and issues regarding the direct-to-consumer (DTC) genomic testing business and market.

It may be interesting to look at reactions to key events of the DTC GT market within the Twitter community. This will require Twitter’s Premium API to access records from the past and an evaluation of specific periods that are worth analysing. Annual Black Friday and Cyber Monday may also be suitable for trend analysis.

References

Twitter as a Tool for Health Research: A Systematic Review

Quantifying Mental Health Signals in Twitter

Characterisation of Mental Health Conditions in Social Media using Informed Deep Learning

…and much more

To be updated.