A video game-inspired approach offers a new way to visualize limb development

From a single cell to the whole organism, embryonic development is a continuous and constant process of change. However, understanding of this process can be limited by the fact that this process can only be observed as “snapshots,” particularly finer molecular details such as gene expression patterns.

Currently, scientists in the research group of James Sharpe of EMBL Barcelona have found ways to visualize the continuous evolution of gene expression patterns in development by applying techniques commonly used by video game developers. This study was recently published in Journal Development.

Here, Dr. Laura Avigno Esteban students and the first author of the study explains how new discoveries were born.

“One of the biggest challenges in developmental biology is the inability to observe internal embryo development in real time. In some animals, such as fish and amphibians, it occurs outside the parents’ bodies, allowing researchers to directly shoot films of the process, but in mammals like mice, organ formation only occurs within the uterus.

“This will lead researchers to rely on collecting embryo snapshots at different stages if they want to study development. This approach is inherently limited, but requires a large number of samples.

“Studying gene expression patterns is even more complicated. Unlike anatomical structures (such as bone or specific tissue), gene expression patterns are not clearly linked to the physical location of Ebrio.

“What we really need is a continuous explanation of these dynamics, allowing us to visualize and even model gene expression patterns in embryos that are developing over time in meaningful ways.

“This is exactly what was established to achieve in Sharp Group studying the development of limbs in mice. Our study uses previously collected datasets from various experiments to reconstruct a smooth, continuous timeline of gene expression patterns and fill in the gaps in previous fragmented descriptions of embryonic development.

“To do this, we have developed a new method for ‘interpolating’ gene expression data. That is, we estimated missing data points for gaps between measurements. Connect the dots on the graph and draw lines to bridge the gaps and complete the data.

“Our method focuses on simpler and more effective strategies, instead of trying to directly spread the whole tissue pattern. We interpolate gene expression in each limb of each moving tissue into a single representation of the developing limb.

“We encountered the important challenge during the development of this method: choosing the right interpolation technique. Two commonly used interpolation methods (linear and polynomial) have their own limitations. Linear interpolation that links dots to straight lines is in the order of unrealistic shapes. Flexibility.

“The breakthrough came from an unexpected source, a video game. One night, while playing, I realized that game developers would use ‘something’ to create smooth and flexible interpolation of camera movements and animations.

“This method provided exactly what we needed: an arbitrary and smooth interpolation method. By adopting B-splines, we were able to create a method that seamlessly integrates various sources of imaging data to generate a smooth reconstruction of time and space gene expression patterns.

“To test our method, we applied it to Sox9, a gene known to mark future skeletal cells during limb development. The results were promising. We also demonstrated that this method works even for genes with very complex patterns, but we tested it in other developmental processes, such as other developmental processes, such as developmental processes.

“In the end, we are extremely proud of this work. Our method addresses one of the most frustrating aspects of developmental biology: the inability to dynamically visualize gene expression. More importantly, we achieved this without the need for additional data collection.”