The evolution of machine learning and artificial intelligence is changing the way the contemporary lab looks and functions. Researchers today are discovering that breakthroughs can happen not only at the bench, but on the desktop.
Noam Auslander, Ph.D., assistant professor in the Molecular & Cellular Oncogenesis Program of Wistar’s Ellen and Ronald Caplan Cancer Center, conducts her research at the intersection of computer science and biomedical science. The interdisciplinary nature of Wistar provides fertile ground for her innovative lab to flourish and tackle research questions from a multifaceted and collaborative perspective. She uses advanced techniques to investigate genetic factors underpinning cancer evolution and viruses to improve diagnostics and therapeutics. As a computational scientist, Auslander applies the power of advanced computational platforms – artificial intelligence (AI) and machine learning (ML) software – to very intricate and complex biomedical data.
Wistar had a conversation with Auslander to find out more about how computer science impacts biomedical research on cancer and viral diseases and the evolution of the next generation research lab.
Can you define artificial intelligence and machine learning?
These are fields in computer science that involve algorithms allowing learning from a set of examples. This can be for instance, learning to make decisions based on data, or to transform one data into another form.
What is unique about using computer science and AI approaches in biomedical research? How far along is the field in biomedical science?
There are some biomedicine areas where AI approaches are very advanced, such as for prediction of protein structure or radiology. However, in other domains, such as for drug identification or precision medicine, it is still in its infancy. Some reasons for these differences are (1) how much data is available to build AI models (2) how much effort is invested to address a particular research question and (3) how well defined the problem is in terms of data and goals.
Why is using advanced computational methods important to biomedical research? What types of knowledge and data can be gathered and analyzed using AI?
We have a huge amount of data that is available to us, and datasets are being generated every day. Within these, there is hidden information that can benefit biomedicine if uncovered, and the only way to do this realistically is by applying computational approaches and improving methodologies that can harness these datasets.
What do you anticipate for the future of computer science techniques in a research space? What does innovation look like in your field?
We need advancements in algorithms and implementations that will make the most use of continuous improvements in computing power.
Just as libraries have dramatically changed with the proliferation of computers and tablets – do you see the lab of the future morphing more into computers and software and fewer test tubes and high-end imaging equipment?
It is already happening – almost every graduate program today is teaching some basic coding skills, and many wet-lab biologists today hire computational staff. This trend will probably continue.
What is one thing you wish people knew about harnessing the power of artificial intelligence for research?
I wish people would be careful and acquire appropriate training before using such tools in their research.
What impact do you hope your work will have on cancer and infectious disease?
We are searching for unknown organisms that are associated with cancers and other diseases. I hope that we will find viruses and bacteria that play a role in these diseases and have not been discovered previously.
How is your day-to-day schedule structured?
I oversee the projects in my lab and some ongoing projects with collaborators. There is a lot of processing data and making sure things are done and defined correctly. For scientific literature, I need to keep track of many fields, including cancer research, viruses and infectious diseases, and new computational advancements ¬– and that’s not easy.
What is your favorite part of work and Wistar?
The best part in work is when we find something that no one found before. Wistar is great because people here are very supportive and collaborative, which allows us to move forward, validate new findings, and apply our methods to other domains.
Do you collaborate a lot and how does your lab work together with other Wistar labs?
We are very collaborative, and we have different types of collaborations. Trainees in my lab decide if they work on collaborative projects and to what extent, and this motivates the ‘how’ my lab works with others.
What advice do you have for those interested in pursuing a career like yours?
Start coding early in life and find programs that allow comprehensive training in computer science and biology.