Defense Season Wrapped: What Makes a Thesis Defense Great?

If you’re reading this, chances are you’re gearing up for your thesis defense. That’s a huge milestone, congratulations! By now, you’ve probably talked about your work in all kinds of settings: conference presentations, lab meetings, even those slightly chaotic attempts to explain your research to family over the holidays. 

From what we’ve seen, the thesis defense is a really special moment where all those different conversations come together. It’s also one of the few times you’ll have such a diverse audience in the same room: your labmates, committee, friends, and family. That mix makes it both exciting and a little tricky. The challenge is finding a way to share the heart of your research with people who know it inside out, while also making it meaningful to those who don’t — all in 45 minutes. So how do you do that?

In this post, we share a few things we’ve learned from watching dozens of great thesis defenses over the last couple years, focusing on two of our very own BE Comm Lab fellows, who have generously shared their slide decks with us. We will break down what we think made their defenses accessible, memorable, and impactful, taking note of strategies we want to incorporate into our own defense preparation. Hopefully, this will also give you a helpful starting point as you craft your own!

The Defenses

Caleb Perez: “Deconstructing the CAR signaling landscape to identify drivers of engineered T cell function,” Birnbaum Lab

In his PhD, Caleb developed a scRNA-seq platform to systematically characterize signaling variants of chimeric antigen receptor (CAR) molecules. CAR T cells are a class of living drugs in which a patient’s immune cells are reprogrammed to attack cancers, leading to transformative responses in many patients. Although the intracellular signaling domains (ICDs) of CARs play an integral role in driving the anti-tumor response, the mapping between ICD composition and associated phenotypic output was not well characterized. Caleb used his platform to map different CAR ICDs to functional outcomes, identifying beneficial signatures in different cancer contexts. His work has broad implications for the design of next-generation signaling architectures optimized for cancer and beyond.

To learn more, check out:

Perez, C. R., et al. Library-based single-cell analysis of CAR signaling reveals drivers of in vivo persistence. Cell Systems, 16(5), 101260 (2025). https://doi.org/10.1016/j.cels.2025.101260.

Viraat Goel: “Deeply resolving chromatin microarchitecture and mechanisms of 3D genome organization,” Hansen Lab

Viraat’s PhD was focused on engineering a new approach for high-resolution mapping of chromatin organization within the 3D genome, a major determinant of mammalian gene regulation. While several chromosome conformation capture (3C) techniques had been developed in recent decades for this purpose, these existing methods struggled to resolve enhancer-promoter interactions and were impractical for targeted study of genes of interest. In building a new 3C method known as Region Capture Micro-C (RCMC), Viraat created a new tool for deeply mapping chromatin architecture and also applied it to discover new structures and mechanisms governing enhancer-promoter regulation. His thesis work is now being extended by his collaborators to link genomic structure to function in a variety of disease contexts.

To learn more, check out: 

Goel, V. Y., Huseyin, M. K., & Hansen, A. S. Region capture micro-C reveals coalescence of enhancers and promoters into nested microcompartments. Nature Genetics, 55(6), 1048–1056 (2023). https://doi.org/10.1038/s41588-023-01391-1.  

Goel, V. Y. et al. Dynamics of microcompartment formation at the mitosis-to-G1 transition. Nature Structural & Molecular Biology 32, 2614–2627 (2025). https://doi.org/10.1038/s41594-025-01687-2. 

Finally, as a quick heads-up before we jump in, this post will reference  several slides from Caleb and Viraat’s defenses. If you’re interested in looking at any of the content more closely, you can find PDF files of the selected slides linked at the bottom of this post.

Why Your Work Matters: Speaking to the Whole Room

So where do you start? 

Previous defenders have told us that one of the most challenging aspects of the defense is the background. While most of us are used to recycling slides for various presentations with only a few tweaks, the newly minted PhDs advised us to consider making background slides from scratch. A core part of earning a PhD is being able to clearly explain the unique contribution you’ve made to your field. To do that well, you need to show where the field was when you started. That context matters — not just for fellow scientists, but for anyone trying to understand the significance of your work. This is an opportunity to thoughtfully include those who aren’t accustomed to scientific presentations. Zoom out: what is the essence of your project?

On a very high level, Caleb’s PhD was about developing tools to understand biology in order to improve existing cancer therapies. On the other hand, Viraat developed tools to fill a large knowledge gap in fundamental biology. This difference in their work influenced where they began their introduction. 

Here are their slides in action:  

Caleb Caleb started with introducing the concept of using the immune system as a cancer therapy. This approach was really effective for helping his non-STEM friends and family appreciate the problem he’s trying to tackle, without overwhelming them with all the complexities. After the first three slides, Caleb tailored the rest of his background section to the scientists in the audience, focusing on specific concepts in immunotherapy.  After explaining how CAR-T therapy works, he discusses what aspects can be engineered to make it more effective. He ends the background section by highlighting a knowledge gap in therapy design.

Viraat Viraat started by explaining genes and gene regulation, which are fundamental concepts in biology. Understanding the importance of the genome and how elements like enhancers regulate it are important prerequisites to appreciating the large knowledge gap in the field of 3D genome organization that motivated his thesis. After presenting this basic but essential background, Viraat is able to move into a case study about what can go wrong in the process. He does this by walking us through a real-world context that the audience is familiar with: leukemia. Only after presenting this background information does he then move into discussing why the knowledge gap exists, and how he intends to develop technology to fill it.

Caleb and Viraat also both took a few minutes at the end of their presentations to re-emphasize the significance of their work. Unlike a normal conference talk, the defense is a unique opportunity to not only share research findings, but also highlight the exact niche that your thesis fills, give an honest appraisal of the limitations of your work, and provide a thoughtful outlook on where the field can go from here.

Caleb Caleb created an outlook slide that highlights the contributions of his tool and the possibilities that stem from it. He then showed two extensions of his work, one an ongoing collaboration and the other a parallel project in his lab.

Viraat Viraat spotlighted a couple ongoing projects with collaborators who are applying RCMC to answer new biological questions. He then ends by reflecting on the questions his technique is not suited to address, as well as future improvements and application areas.

Tour Guiding Your PhD

Something that impressed us about both Caleb and Viraat’s defenses is that it was clear they had a huge body of work, but they still managed to distill the most salient results and takeaways. They both organized their presentations with a roadmap that was clearly laid out in the beginning, then revisited at the end.

Caleb Caleb’s first roadmap slide laid out the vision and high-level goals for the tool he created, with each section corresponding to a component of the tool that needed to be developed. When he returned to this slide at the end of his talk, he summarized the capabilities of the technology and filled in the specific biological insights gained from applying his tool.

Viraat Viraat’s roadmap was an outline of the whole presentation, showing the audience a table of contents for the motivation, development, and application of his technology. He returned to this slide several times as he moved through the presentation, keeping audience members focused on the bigger picture. At the end, he even brought in his signature humor by translating each section into a haiku.

When we are thinking about making our own summary slides, it does feel a bit daunting to figure out what are the main landmarks along the tour, and what order they should be presented in. This involves choosing a narrative that is the easiest to understand, rather than one that is strictly chronological. It’s also worth noting that the amount of time you spend on each part may not necessarily be reflective of the time spent doing it. We’ve always enjoyed when people say “you may not realize it, but this slide alone represents three years of the PhD.” It is a helpful reminder to the younger grad students in the audience who might feel stuck in the hard places themselves.

Many theses don’t follow a linear narrative and work better as discrete chapters. For instance, if you implemented a tool for different applications, explored two distinct biological hypotheses, or ended up pursuing entirely unrelated projects, it is worthwhile to think about interspersing your background information throughout the presentation. That is, can you start with just the background that is minimally necessary for your overall thesis, and then save the specifics for when it becomes important? 

Viraat walked us through a useful exercise where he peeled back the layers of background necessary for understanding his first project (developing RCMC). The exercise involves working backwards by asking a series of “why” questions until you reach a fundamental fact that is known to everyone in the room. Presenting the answers to these questions in the reserve order then builds a logical flow of information that you can use to tour-guide your background section. In practice, you might decide to condense some intermediate steps for brevity. 

Boosting Your Signal to Noise Ratio

A concept that we frequently invoke in the Comm Lab is your “signal to noise ratio.” While it might be tempting to emulate someone else’s aesthetics in your own slide making, we encourage you to instead think about the principles of clarity over beauty. Beautiful slides are not always the most interpretable (though Caleb and Viraat managed to do both).

Boosting your visual signal to noise ratio usually means remaking existing figures that you might have from publications or more technical audiences. For example, let’s look at Figure 3 from Viraat’s paper, versus the version that ended up in his defense slides:

Figure from Goel et al. (2023)

Figure from Defense Slides

Here are some of the modifications he made: 

• He removed all of the additional information about the features of that genomic region, which was not an essential part of the story on that slide.
• He highlighted key features of the map using a larger font size and different colors

Importantly, Viraat also identified that people have trouble reading the information-dense contact maps in panels A and B, which are the bread and butter of his field. So, he started his data walkthrough by teaching the audience how to actually read and interpret these figures, before diving into the results in the figure above. We thought this was a particularly effective use of animations to slowly build up to the full complexity of these contact maps, with each click introducing a digestible piece of information:

Thanks to this tutorial, we were able to focus on the conclusions of the contact maps in the rest of the presentation rather than struggling to take in all that information at once. 

Caleb’s presentation also had some great examples of how to maximize visual signal to noise ratio:

Caleb used color to complement the scientific message. He identified a few different clusters of candidate CAR molecules, assigning each of them a color in the first slide that they appear. He returns to this same color scheme at several later points in his talk, as we focus specifically on the k7 candidate CAR shown in cyan. We also love how his slides don’t feel crowded with information! There is plenty of white space.

Another thing to consider when presenting to a non-technical audience is the importance of simple, digestible (but thoughtful!) metaphors and analogies (and even tasteful memes) to make complex information more accessible. Here are some ways they communicated key technical concepts in non-technical ways:

Caleb Some examples: a magnifying glass to illustrate immune surveillance; a toolbox to convey the purpose of exploring different CAR intracellular domains (ICDs); and interconnected arrows to illustrate input vs. output complexity.

Viraat Using humor can be an effective strategy to focus in on what you want your audience to take away.  For instance, Viraat used Kendrick Lamar as a unit of measurement to help us visualize the length of DNA in the human body: Entirely coincidentally, Viraat also used this beloved Drake meme format to underscore the superiority of his suite of technologies:

In summary, good visual signal to noise ratio looks like:

• Figures that only contain the essentials
• Strategic use of font size and bolding to emphasize most important takeaways
• Well-paced animations to present digestible chunks of information
• Consistent colors that have clear meaning

In addition to visual signal to noise ratio, Caleb and Viraat were both extremely good at boosting their verbal signal to noise ratio. Some strategies we noticed include:

• Repeating basic but useful definitions
• Simplifying technical synonyms; for instance, the terms “checkpoint blockade” and “checkpoint inhibitors” refer to the same immunotherapy and are often used interchangeably. Picking one term and sticking with it throughout the presentation can improve clarity.
• Syncing your verbal cues to visual cues when walking through a slide

Finally, one last thing to boost your signal to noise ratio (as well as address speaking to the whole room) is considering the physical space and the AV equipment for your defense. Some of the nonverbal components of a good defense mirror those of any good presentation, but they can sometimes get overlooked amidst all the craziness leading up to a defense. Thinking through the logistics ahead of time can help you feel more confident and in control during the defense itself and avoid the stress of last-minute panic if something isn’t working.

From chatting with Viraat and watching defenses throughout the years, here are some of our suggestions:

• Scope out your defense room ahead of time: know where the podium is, how you’ll face the room, which screen(s) you’ll use and what information they’ll show, whether or not you’ll have speaker notes, how you want to set the lighting in the room (e.g., dimmed if you have a lot of black slides, brighter if not). Practice giving the defense up at the front of a similar room, being mindful not to read off the slide and maximizing eye contact with the audience.
• Figure out how you’ll manage and connect with the in-person and virtual audiences. Test out your Zoom room link/settings and designate a friend/labmate as a co-host who can help moderate Q&A with the virtual audience, as well as mute the one person who inevitably joins with their audio on partway through. Think about how virtual audiences will see things differently than physical ones, too — for instance, avoid physically pointing at the screen and instead use a virtual pointer on the screen so that all audiences can follow.

You Do You (and Thank Them Too)

While the science is the highlight, one of our favorite parts of defense season is getting to see all the people behind the science. We love seeing people’s personalities shine when they get to the more personal acknowledgments segment of their defense. Hearing them thank their mentors, labmates, and loved ones is a wonderful reminder that although a PhD is marketed as a very solo adventure, in BE, it is more often than not a team effort. We’ve seen several ways to do this — short and sweet, or listing every single BE community member who has ever existed — and either strategy is reflective of the person behind the science.

In the end, a PhD is also deeply personal. This is your work in a way that almost nothing else ever is. We hope the principles we laid out in this post will allow you to honor both the substance of your work and your individuality.

Example Slides

Here are a selected subset of Caleb’s and Viraat’s slides:

Caleb’s Slides

Viraat’s Slides

Blog post written by BECL Fellows Ellen Kan, Pallavi Balivada, Caleb Perez, Viraat Goel, with editing help from BECL Manager Han Xu.

Posted Feb 2026.