From the outside, scientific breakthroughs can seem straightforward; however, beneath the surface lies the iterative process of testing ideas, revising assumptions, and confronting unexpected complexity. We asked early-career researchers, from cancer and computational biology to microbiology, to share their thoughts on the matter: What’s a common misconception about your area of research?
Cancer Biology: The Many Forces That Shape Tumors
Kasturika Shankar, Biochemist, University at Buffalo
Many people assume that because we have a human papillomavirus (HPV) vaccine, HPV-driven cancers are essentially a solved problem. While the vaccine is genuinely one of modern medicine's success stories, global uptake remains low, particularly in lower-income countries where access is limited and in populations where vaccine hesitancy has taken hold. The reality is that HPV-positive head and neck cancer is already on the rise, and projections suggest it could become one of the most common cancers in the coming decades. Patients diagnosed with these cancers today were not protected by the vaccine, and they need better treatments now. That's why understanding the biology of these tumors deeply and mechanistically remains urgent, even in a world where a preventive tool exists.
Fariha Imtiaz, Biologist, Virginia Commonwealth University Massey Comprehensive Cancer Center
Awareness of advanced glycation end-products (AGEs) remains limited, despite their formation both endogenously and through environmental exposures, poor diet, limited healthcare access, and socioeconomic disadvantage, reflecting a broader biosocial synergy. These molecules are compounds formed when proteins or lipids become glycated after exposure to sugars. They are often perceived as either harmless or relevant only to conditions such as diabetes. However, AGEs can modulate inflammation and the cellular microenvironment, thereby promoting cancer progression. Addressing these misconceptions is essential, as it reframes cancer not solely as a genetic disease but as a condition also influenced by long-term metabolic and environmental factors.
Tarang Gaur, Cancer Biologist, Albert Einstein College of Medicine
The biggest misconception is that findings from standard laboratory cell lines directly translate to patients. My work with primary patient samples and patient-derived xenograft models repeatedly shows this isn't true. Cancer cells grown in a dish for decades behave fundamentally differently from cells freshly isolated from a patient's bone marrow. Perhaps most frustratingly, people assume that identifying a good drug target automatically means a drug will work. Biology is rarely that cooperative. The gap between a promising target and a clinically effective therapy is enormously filled with resistance mechanisms, toxicity challenges, and biological complexity that no textbook fully captures. That gap is exactly where my research lives.
Cellular and Molecular Biology: Beyond “Junk” DNA and Lab-Grown Organs
Kaushik Chanda, Molecular Biologist, The Herbert Wertheim UF Scripps Institute for Biomedical Innovation & Technology
A common misconception about non-coding RNA research is that these molecules are “junk” or biologically unimportant simply because they do not encode proteins. In reality, non-coding RNAs, especially circular RNAs, are increasingly recognized as critical regulators of cellular function, particularly in the brain where they are highly abundant and evolutionarily conserved.
Róża Przanowska will join Loyola University Chicago as an assistant professor in the fall of 2026.
Amanda Maglione
Róża Przanowska, Cancer Biologist, University of Virginia
For many years, non-protein-coding regions of the genome were dismissed as “junk DNA,” and remnants of that perception persist. Some people assume that long non-coding RNAs (lncRNAs) are too weakly expressed or poorly conserved to have meaningful biological functions. However, decades of research have demonstrated that lncRNAs regulate critical cellular processes, and many have important roles in development and disease. In cancer biology especially, we are increasingly recognizing that molecules once overlooked may hold important clues to diagnosis and treatment.
Khyati Raina, Biologist, Albert Einstein College of Medicine
Many assume that stem cell-based therapies are already ready to be used widely in patients. There’s been a lot of exciting progress, which makes it seem like we’re almost there; but most of this work is still in the research stage. We can grow organ-like structures and specific cells in the lab, which is amazing, but making sure they work as intended and are safe in humans requires a lot more work.
Ritu Ramamurthy, Biomedical Engineer, Wake Forest University
People often think that organoids are tiny, fully functional organs. They are research models designed to capture specific features of human biology. They are not perfect replicas, but they can still provide insights that are difficult to obtain from simpler culture systems or animal models. Another misconception is that these models are meant to completely replace animal research. I see them more as an important additional tool that can help make research more human-relevant and improve translation to the clinic.
Microbiology and Immunology: Pathogen Interactions and the Immune System
Aditya Gupta, Parasitologist, United States Department of Agriculture
A common misconception is that parasites in meat are rare or only a concern in regions with limited food safety infrastructure. Many parasitic infections are underreported because they can be asymptomatic or difficult to detect using routine inspection methods. Another misunderstanding is that visual inspection alone is sufficient to identify contaminated meat. Many parasites, including Sarcocystis, form microscopic cysts that require molecular tools for accurate detection. Increasing awareness of these limitations is important for improving surveillance and food safety.
Gundappa Saha, Immunologist, University of Pennsylvania
Some people assume that the innate immune system is simple, short-lived, and lacks memory, but it can undergo long-lasting changes that shape how it responds in the future. Another misconception is that chronic diseases develop independently of each other. Increasing evidence suggests that many of these conditions are interconnected through shared immune pathways, meaning that treating them in isolation may overlook the underlying cause.
Abhilash Nair, Microbiologist, Columbia University
Many think that advances in microbiome research have made the field simpler. In reality, they have revealed just how complex and vast it truly is. As our tools improve, so do the questions, uncovering layers that remain largely unexplored. In a rapidly changing world, understanding these systems requires deeper insight, as the challenges are becoming more intricate rather than easier to solve.
Computational Biology: The Limits of Prediction Models
Kun Bu will join Pepperdine University as an assistant professor in the fall of 2026.
Kun Bu
Kun Bu, Data Scientist, University of South Florida
A common misconception is that artificial intelligence and large-scale data analysis can automatically uncover objective truths from data. Data are often incomplete, noisy, and shaped by how they were collected. Even the most advanced algorithms can inherit biases or identify patterns that do not reflect true underlying relationships.
Mary Cundiff, Computational Biologist, University of Pittsburgh
A common misconception is that more complex machine learning models automatically lead to better biological insight. In reality, highly complex models can make it harder to understand what drives the results. In biology, interpretability is essential. If we cannot explain why a model identifies a pattern, it becomes difficult to trust or act on those findings. My work emphasizes that simpler, interpretable frameworks can often provide deeper insight into biological mechanisms than black-box approaches.
Anupam Ojha, Computational Biophysicist, Flatiron Institute
A widespread assumption outside structural biology is that recent advances in protein structure prediction, particularly AlphaFold, have effectively solved the structure problem. They have not. These models predict, with remarkable accuracy, a single low-energy conformation per sequence. They do not predict the populations of states a protein visits, nor how those populations shift under mutation, ligand binding, or environmental changes. For most therapeutically relevant proteins, function depends on differences between conformations, not on any single conformation. The next frontier is predicting ensembles, with experimentally calibrated weights, rather than structures.
Responses have been edited for length and clarity.
