In this interview, we speak to Victor Wong, Chief Scientific Officer at Core Life Analytics, about their StratoMineRTM product and how it is helping researchers to rapidly process their data.
Please could you introduce yourself and tell us about your journey to Core Life Analytics?
My name is Victor, and I started my scientific career at the University of Toronto, where I did my PhD in Physiology. My focus then was on metabolic disorders, with an emphasis on drug targets and drug discovery. I then worked as a postdoc in neuroscience at the UC Davis and Weill Cornell Medical Center. At the latter institution, I was exposed to high throughput and high content imaging, using compound screens for drug discovery.
My naivete initially gave me a false impression that automation would significantly accelerate my projects and publications, but that was simply not the case. Data analysis was the biggest challenge; the amount of data coming from my projects was beyond my knowledge to even know where to start. I tried grasping some mastery of programming, but nothing was ever robust or reproducible.
I joined Core Life Analytics simply because they are the solution to the data problem that I had. Also, our scientific philosophies align incredibly well: to provide robust and transparent data analytics tools that allow scientists to analyze their data quickly and paint a holistic picture of their experiments. Moreover, and more importantly, to educate scientists about the good practices in data science.
What are Core Life Analytics' primary aims, and how does it fit into the broader field of biological and life sciences?
At Core Life Analytics, we are on a mission to democratize data science: we help biologists to analyze their complex phenotypic data independently.
High-content or phenotypic screening is a powerful tool for drug discovery. Using advanced microscopes and image analysis software, scientists can translate microscopic images to hundreds or thousands of measurements of a cell's morphology, such as size, intensity, and shape. These so-called features describe and quantify a cell's phenotype, allowing researchers to assess a compound's effect carefully.
Techniques like these fit in a movement towards more data-driven approaches: instead of focusing on measurements you know to be involved in the processes you study, measure them all, and use statistics to determine what is interesting.
Image Credit: Gorodenkoff/Shutterstock.com
The rise of data analytics and bioinformatics is prevalent across all areas of biological and life sciences, yet many still face challenges in integrating it into their workflow. What are the main barriers that limit the use of advanced data analysis software in these sectors?
Most scientists struggle to make use of these datasets. Our founders, David Egan and Wienand Omta, witnessed this first-hand at the UMC Utrecht Cell Screening Core; their clients either had to learn data science and coding skills to analyze their data or ask a data scientist to do it for them. In both cases, data goes underutilized. Often, only a handful of known measurements are analyzed, leaving hundreds or thousands of potentially useful ones behind.
Why should biologists endeavor to use advanced data analysis platforms, and how can this help to catalyze innovation in sectors like drug discovery?
Giving biologists the tools to perform their analyses independently greatly improves the speed with which discoveries are made. Firstly, biologists no longer need to wait for busy data scientists or learn to code but can simply run their data through the analysis platform.
Secondly, this enables the person who knows the experiments best - who designed and executed them - to explore the data and make decisions for future experiments accordingly. This is not only important for the final analysis of an experiment; being able to quickly run these analyses in the early, experimental stages of a study helps assess the quality of your model or assay and identify problems early on.
Don't forget about the data scientists; when biologists perform these relatively routine analyses themselves, they have time for the more exciting stuff, such as advanced AI and multi-omics.
StratoMineRTM is Core Life Analytics' primary product that aims to help researchers rapidly process their data. Could you discuss the background behind this product and how users can integrate it into their workflow?
When David Egan and Wienand Omta realized at the UMC Utrecht that the need for accessible data analysis tools was widespread, they decided to develop something that could be used by any biologist dealing with this type of data. Regardless of the hardware or software they use or their data science skills. Simply upload your numerical data, and StratoMineR guides you through a best-practice workflow for phenotypic data.
Starting with the more basic steps, such as finding relevant features, performing quality control, and normalization and scaling of your data, to more advanced steps, like data reduction, to eventually comparing and grouping phenotypes to determine a compound's mechanism of action.
How does StratoMineRTM compare to existing platforms currently available? Are there any components that end-users would find especially interesting?
What differentiates our approach from other tools is that it is intuitive and decision-supportive. StratoMineR's guided workflow ensures no step is missed and offers suggestions using AI where possible. This way, any biologist can follow a best-practice analysis workflow for multiparametric data, understand and explore it. And start doing so early in the experimental phase of a project.
Core Life Analytics recently attended ELRIG Drug Discovery, Europe's largest meeting that brings together industrial professionals across life sciences. What are the benefits of attending such events to discuss and demonstrate products in person?
ELRIG Drug Discovery was a great meeting with an excellent scientific program. We always enjoy meetings like these, as they are a perfect opportunity to keep up with the latest developments in the field. Most importantly, we can talk to scientists from many different backgrounds and learn about their perspectives and challenges.
Many advancements are being made in data science technologies, with all industries reaping the benefits. How do you expect the relationship between data science and the life sciences sector to change over the next ten years?
As mentioned earlier, the interest in data-driven drug discovery is growing. A great illustration of this is the JUMP-CP Consortium, which has generated a database of phenotypic data from cells responding to 140.000 different genetic perturbations and small molecules.
The potential of this public resource is obvious, but it raises the question: how can researchers outside of the consortium leverage a large and complex dataset? This and the ever-increasing data complexity further highlight the need for accessible tools. We already see that AI-based analysis tools are becoming more and more mainstream, Machine Learning (ML) and Deep Learning (DL) are emerging, and discussions on more advanced integrations, such as multi-omics, are ongoing, developing into a new research area.
Image Credit: paulista/Shutterstock.com
What do the next few years look like for Core Life Analytics? Are there any innovations you are striving towards?
Over the coming years, we hope to tackle some other bottlenecks in high-content screening. One of our ambitions is to, in addition to the numeric data, move the images to the cloud. This will resolve many people's storage issues and allow us to use massively parallel cloud computing for image analysis, dramatically reducing analysis time.
Where can our readers go to stay up to date with the company's activities?
They can follow us on LinkedIn or visit our website.
Please provide links to any materials that may be relevant to our audience.
On Nov 15th, we will host a webinar: Get Ready for JUMP-CP!
More information on the JUMP-CP Consortium can be found on their website.
More information on StratoMineR for high content data can be found in our brochure.
About Victor Wong
As a CSO, Victor Wong's responsibilities are to establish and communicate the scientific validity and utility of research products developed by Core Life Analytics. He interacts with scientific and customer communities regarding the capabilities and scientific findings from our company. He also manages with other CxOs to the overall management of the products and the team.
Victor Wong received his Ph.D. at the University of Toronto. He was a Canadian Institute of Health Research Fellow, and received a number of grants during his postdoctoral training at the Burke Institute of Weill Cornell Medicine. His scientific motivation is driven by his disability; he is profoundly deaf and since then, his scientific journey led him through a number of therapeutic fields, with a focus on target and drug discovery to find novel treatments for a number of diseases spanning metabolism, oncology, neurodegeneration and hearing loss.