What if the price of machines that think is people who don't (George Dyson, Turing's Cathedral).
Tumors are often said to reuse genetic programs from early tissue formation, but similarity to those programs does not mean cancer cells follow the same trajectories. We developed a quantitative framework to separate these ideas in pancreatic ductal adenocarcinoma. Variation tied to normal epithelial differentiation provided a useful coordinate system for locating malignant cell states, yet differences between those states showed weak alignment with canonical differentiation directions and included structured, cancer-adaptive programs. Thus, pancreatic cancer retains aspects of normal cell identity while diversifying along additional malignant axes. This distinction clarifies how insights from normal tissue programs should be applied to cancer plasticity and provides a general method for comparing disease states with normal cellular programs. Read more here.
Why do lineages repeatedly evolve extensive somatic investment, despite the fact that somatic cells are largely excluded from future generations? Classical discussions of the soma-germline distinction often define germline and soma by anatomy, developmental origin, reproductive function, or cell fate. We propose a complementary population-genetic perspective in which germline identity is defined by the capacity of genetic material to contribute to subsequent generations, and soma corresponds to lineages whose genetic variation is excluded, restricted, or only conditionally admitted into evolutionary continuity. Have a look!
Housekeeping genes are the basic maintenance genes that cells need to stay alive, so they tend to be active in most cells, unlike specialized genes that define what makes a brain cell different from a muscle cell. However, most studies define housekeeping genes using adult tissues only, which misses how gene activity changes during development. Using zebrafish embryos and tracking which genes are active across different stages of embryonic growth, we evaluated a new definition of these genes based on their consistent throughout development. Results not always coincide with those found with the traditional approach here.
Development is a contextual process where the state of a cell is changing along the organismal maturation. We can now follow this process at the single cell level and quantify the activity of genes on each cell. We want then to understand how the context of a cell influences the activity of genes, and how this can be captured by large language models. Some first efforts with Zebrafish data here.
It is commonly argued that genes that are active in similar ways during development tend to produce similar effects on an organism’s traits. But is this really the case? To test this idea, we are analyzing gene activity in different cell types during the development of the nematode Caenorhabditis elegans, using a detailed single-cell gene expression atlas. We then compare these patterns to a curated database of phenotypes, observable traits or behaviors linked to specific genes. This is what we found.
We aim to know how well genetic data can predict complex traits, like behavior or disease risk. Traditional models assume gene effects add up simply, but biology is often nonlinear and network-driven. Using simulations of gene regulatory networks, we explore when these assumptions hold, and when they break down. Our work helps clarify the limits of genetic prediction tools like polygenic scores. Here, you can find details.
We explore the idea that cells and tissues manage protein production to minimize energy costs. By analyzing highly expressed proteins across human cells, tissues, and tumors, we show that this metabolic efficiency principle helps explain which proteins are favored, and when this pattern breaks down. This approach offers a tractable way to understand the energetic logic behind both normal cellular function and disease. Want to read more? Here.
Biology involves layered mappings -from genes to traits to fitness- all influenced by variation. This variation can be masked, redirected, or amplified, shaping robustness, plasticity, and evolution. I propose here that general design principles govern how variation moves through these layers, a concept I call multi-map variation. Identifying these principles may help us better understand and predict evolutionary change.