New computer tool predicts cell fate and genetic disruption | MIT News
Imagine a ball thrown into the air: it curves up, then down, tracing an arc to a point on the ground some distance away. The trajectory of the ball can be described by a simple mathematical equation, and if you know the equation, you can determine where the ball will land. Biological systems tend to be harder to predict, but MIT biology professor Jonathan Weissman, postdoc Xiaojie Qiu, and collaborators at the University of Pittsburgh School of Medicine are working to make the path taken by cells as predictable as the arc of a bullet. Rather than looking at how cells move through space, they consider how cells change over time.
Weissman, Qiu and collaborators Jianhua Xing, a professor of computational and systems biology at the University of Pittsburgh School of Medicine, and lab graduate student Xing Yan Zhang built a machine learning framework that can define the mathematical equations describing a cell’s trajectory from one state to another, such as its development from a stem cell into one of several different types of mature cells. The frame, which they call “dynamo”, can also be used to understand the underlying mechanisms – the specific cocktail of gene activity – driving changes in the cell. Researchers could potentially use this information to manipulate cells to take one route over another, a common goal of biomedical research and regenerative medicine.
Researchers describe the dynamo in one paper published in the journal Cell February 1. They explain the framework’s many analytical capabilities and use it to help understand the mechanisms of human blood cell production, such as why one type of blood cell forms first (appears faster than others).
“Our goal is to move towards a more quantitative version of single-cell biology,” says Qiu. “We want to be able to map how a cell changes in relation to the interaction of regulatory genes as precisely as an astronomer can plot the motion of a planet in relation to gravity, and then we want to understand and be able to control these changes.
How to map the future journey of a cell
Dynamo uses data from many individual cells to find its equations. The main information it needs is how the expression of different genes in a cell changes from moment to moment. Researchers estimate this by looking at changes in the amount of RNA over time, because RNA is a measurable product of gene expression. In the same way that knowing the starting position and speed of a bullet is necessary to understand the arc it will follow, researchers use starting levels of RNAs and how those RNA levels change to predict the path of the cell. However, calculating changes in the amount of RNA from sequencing data from a single cell is difficult because sequencing only measures RNA once. Researchers must then use indices such as RNA being made at the time of sequencing and RNA turnover equations to estimate changes in RNA levels.
Qiu and his colleagues had to improve on previous methods in several ways to get measurements clean enough for the dynamo to work. In particular, they used a recently developed experimental method that marks new RNA to distinguish it from old, and combined it with sophisticated mathematical modeling, to overcome the limitations of old estimation approaches.
The researchers’ next challenge was to move from observing cells at specific times to a continuous picture of how cells change. The difference comes down to going from a map showing only landmarks to a map that shows the uninterrupted landscape, allowing paths to be traced between landmarks. Led by Qiu and Zhang, the group used machine learning to reveal the continuous functions that define these spaces.
“There have been tremendous advances in methods for general profiling of transcriptomes and other ‘-omics’ information with single-cell resolution. The analytical tools to explore this data, however, to date have been descriptive rather than predictive,” says Weissman, who is also a fellow at the Whitehead Institute, a fellow at the Koch Institute for Integrative Cancer Research, and a Howard Hughes Medical Institute researcher. “With a continuous function, you can start doing things that weren’t possible with just precisely sampled cells at different states. For example, you can ask: if I modify a transcription factor, how will it modify the expression of other genes? »
Dynamo can visualize these functions by transforming them into mathematical maps. The terrain of each map is determined by factors such as the relative expression of key genes. A cell’s starting point on the map is determined by the current dynamics of its gene expression. Once you know where the cell begins, you can trace the path from there to find out where the cell will end.
The researchers confirmed Dynamo’s cell fate predictions by testing it against cloned cells — cells that share the same genetics and ancestry. One of the two nearly identical clones would be sequenced while the other clone would differentiate. Dynamo’s predictions of what would have happened to each sequenced cell matched what had happened to his clone.
Moving from mathematics to biological understanding and non-trivial predictions
With a continuous function for a cell’s path over time determined, the dynamo can then better understand the underlying biological mechanisms. Calculating derivatives of function provides a wealth of information, for example allowing researchers to determine the functional relationships between genes – if and how they regulate each other. Calculation of acceleration can show that the expression of a gene is rising or falling rapidly even when its current level is low, and can be used to reveal which genes play a key role in determining the fate of a very cell. early in the cell’s trajectory.
The researchers tested their tools on blood cells, which have a large branching tree of differentiation. Along with blood cell expert Vijay Sankaran of Boston Children’s Hospital, the Dana-Farber Cancer Institute, Harvard Medical School and the Broad Institute of MIT and Harvard, and Eric Lander of the Broad Institute and the Department of Biology at MIT, they discovered that the dynamo accurately mapped blood cell differentiation and confirmed a recent finding that one type of blood cell, megakaryocytes, forms earlier than others. Dynamo also discovered the mechanism behind this early differentiation: the gene that drives the differentiation of megakaryocytes, FLI1, can self-activate and therefore is present at relatively high levels early on in progenitor cells. This predisposes the progenitors to first differentiate into megakaryocytes.
The researchers hope that the dynamo could not only help them understand how cells go from one state to another, but also guide researchers in controlling them. To this end, the dynamo includes tools for simulating how cells will change under different manipulations, and a method for finding the most efficient path from one cell state to another. These tools provide a powerful framework for researchers to predict how to optimally reprogram any cell type to another, a fundamental challenge in stem cell biology and regenerative medicine, as well as to generate hypotheses about how d other genetic changes will alter the fate of cells. There are a variety of possible applications.
“If we design a set of equations that can describe how a cell’s genes regulate each other, we can computationally describe how terminally differentiated cells turn into stem cells, or predict how a cancer cell can respond to various drug combinations that would be impossible to test experimentally,” says Xing.
Dynamo goes beyond simple descriptive and statistical analyzes of single cell sequencing data to derive a predictive theory of cell fate transitions. Dynamo’s toolset can provide deep insight into how cells change over time, hopefully making cell trajectories as predictable to researchers as the arc of a bullet, and therefore as easy to modify as modifying a step.