Abstract
Bioscience is becoming increasingly digitalised with new breakthroughs in imaging, protein structure predictions and an active “omics” space all of which produce vast amounts of data and have complex computing requirements. Whilst the possibilities of this space are enormous it presents several challenges: large amounts of fast data storage, GPUs and software that can be difficult to install and maintain. Given the rapid pace of AI expansion research groups are facing two obstacles – keeping up with the domain and a growing expense to process data. All these must be overcome whilst driving research science. The growing area of digital Research Technical Professional (dRTP) can help with these challenges however as this is a new area finding resource can be hard.
Over the past 5 years, the Rosalind Franklin Institute Advanced Research Computing team has built a flexible platform that is by researchers to analyse their data. We have fully integrated laboratory data, provided a suite of installed common software and provided scalable GPU power as well as giving researchers the ability to install their own software and write code they need allowing them to focus on science. This talk covers how we built and maintain the platform and provides some insight into architectures that might be used in other places to achieve similar results.
Biography
Dr Laura Shemilt is Head of Research Software Engineering at the Rosalin Franklin Institute. Her main focus is to collate and curate the data collected at the Franklin into a usable format for analysis and distribution.
She studied physics at Imperial College London before completing a PhD in X-ray Imaging at UCL. She has worked between academia and business in various Data Science and Engineering roles in various exciting companies and fields.