A Computer Scientist Reports On Life In Pharma R&D

January 27, 2020

January 27, 2020 | “I think that pharma R&D is a fascinating place for a computer scientist to work!”

So says Felipe Albrecht, Bioinformatics and Computer Scientist, Pharma Research and Early Development (pRED), at Roche in the city of Penzberg, located in the Greater Munich Area. Albrecht is a data scientist at heart, believing that data drives all science. And his degrees are all in computer science. But when it came to applying that passion, he turned to molecular biology.

Now he works at Roche, constantly learning about pharma, immunology, and biochemistry from his colleagues as he applies data science to analyze and make the most of their work.

On behalf of Bio-IT World, Hannah Loss spoke with Albrecht about his inspirations, challenges, and hopes for working at the intersection of computer science and drug discovery.

Editor’s Note: Hannah Loss, a conference producer at Cambridge Healthtech Institute, is working on the Cloud Computing track at the upcoming Bio-IT World Conference & Expo in Boston, April 21-23. Albrecht will be speaking on the program; their conversation has been edited for length and clarity.

Bio-IT World: What inspires you as a data scientist?

Felipe Albrecht: Data drives science. As a fundamental example, Darwin came up with the theory of evolution after observing many species and sketching a phylogenetic tree, Mendel, the father of genetics, made tables containing his observations after cultivating and crossing different peas. So what they did then, and what many scientists do, is data analysis! A dear colleague of mine here at Roche likes to say, "there is no science without data science." Because what we do here is mainly producing and analyzing data for obtaining knowledge.

Nowadays, due to the vast amount of data, it is necessary to have people with computer science and data science background for supporting the task of handling and analyzing all the data that is being currently generated.


With a background in computer science and software engineering, how did you end up in the pharma R&D world?

Yes, my background is really computer science. My bachelor's and master's, and my PhD is in computer science departments. But I always have a fascination with science, especially molecular biology. During my bachelor's, master's, and Ph.D., I worked on projects related to proteins, DNA sequences, and, lately, epigenetics. Together, I love computers, programming, databases, networks, operational systems and so on. And I always enjoyed learning about how things work. So, what is more interesting than trying to understand how life works?

So, I had the opportunity of joining Roche to work on automation and data handling projects. What I enjoy here is that my work pushes me to learn more about pharma, immunology, and biochemistry. The work pushes me to be able to collaborate with my colleagues in our primary goal that is developing the best medicines for patients. I genuinely believe in the impact that our work here brings to people worldwide. 


What was it like transitioning into working with pharma?  

From the technical side, before joining Roche, I already had some background in molecular biology. However, as I said before, I still have to study pharma, biochemistry, and molecular biology daily to be able to cope with my tasks. It is a bit tough, I have to say, but interesting for a scientist to always have to learn more and more.

From the cultural side, in software companies, it is usual to have terms like a 10x engineer, or a ninja developer, that is an outstanding developer that can solve many problems by himself. But here, the ability to work in a group is much more critical than to be the person that does everything by himself. It is also great because everyone is open to sharing experiences and knowledge.

For example, today I sat with a colleague as he explained the different methods used for extracting DNA sequence. The other day, another colleague was telling me details about mass spectrometry, and in the same conversation I also explained many technical details about programming, software development. In the end there is a cycle of mentoring, I can help them with IT questions that they have, and they can help me with some biological questions.


Do you find that your goals as a bioinformatician and computer science person align with the goals of biochemical scientists? If not, what are the discrepancies?

Our ultimate goal is doing now what patients need next. We must always be aligned in that task. The way of doing this may have some differences. While biochemical scientists are focused on the direct development of medicine—developing antibodies—my task is to provide the best computational tool that they can use for performing such tasks.

Sometimes they need new tools or more features, and my mission in such cases is to prioritize what we are going to deliver to them. Also, currently, we have a deluge of new machine learning and AI tools. They see all these tools at conferences, in the news, in scientific journals, and get very excited about them. That is great! So, it is also my task to keep learning about these tools and find the ones that will bring real benefits to developing medicines.


I see that you're a proponent for "Open Science." What does that mean to you and how do you apply this principle to your current research?

Open Science, for me, is that our knowledge must be shared openly. For universities and research institutes, the papers must be open and available, the data must be open and available. In companies, for IP reasons, we have to be more cautious about sharing our expertise with external actors, but Open Science must also be applied inside the companies.

A good example that I’m working with right now is a project that I will present in the next Bio-IT World. This project's goal is to organize our Mass Spectrometry data for our scientists to reuse the data. Together, we can connect this data with the data produced in other departments, in this way, eliminating the data silos and consequently having open data access.


Do you have any advice for your computer science peers working in the pharma R&D space?

For people in pharma and the R&D space, my primary advice is to participate in the drug development process. We must leave our comfort zone in developing tools and focus on what matters for the patients: that is the development of the best possible medicines, not the best software.

Also, we must never stop learning. In the same way that computational methods are evolving fast, medicine development is also very dynamic. Be open to exploring the pharma world. Learn about pharmaceutical development, molecular biology, and biochemistry.

In the same direction, I advise to spend time together with our biochemical colleagues. Talk with them, listen to them. We can, and we must help to fulfill their computational and data needs.

The final one is for [a computer scientist currently outside of pharma]: be open for new opportunities. There is a lot of interesting computational problems in pharma research and development—from analyzing DNA sequences, protein structures, lab automation, and data handling, processing, and analysis. Also, pharma is a great place for applying machine learning and AI methods. I think that pharma R&D is a fascinating place for a computer scientist to work.


Albrecht can be reached by email at felipe.albrecht@roche.com or on Twitter: @felipealbrecht