06-Jul-2022 - Max-Planck-Institut für molekulare Genetik

How to find marker genes in cell clusters

New method facilitates identification of cell-type specific genes in single-cell data

The thousands of cells in a biological sample are all different and can be analyzed individually, cell by cell. Based on their gene activity, they can be sorted into clusters. But which genes are particularly characteristic of a given cluster, i.e. what are its “marker genes”? A new statistical method called Association Plot facilitates the determination and analysis of these marker genes.

Which genes are specific for a certain cell type, i.e. “mark” their identity? With the increasing size of datasets nowadays, answering this question is often challenging. Often, marker genes are simply genes that have been found in specific cell populations. However, many more genes could be characteristic of a particular cell type but remain undiscovered.

“Association Plots (APL)”, a new statistical method for visualizing gene activity within a cell cluster makes it easier to find its marker genes. The plots compare the activity of genes of a given cluster with all other clusters from the data set. Additionally, they make it easy to see which genes are shared with other clusters.

“Association Plots not only allow us to identify new marker genes. It also works the other way around – we are able to match clusters of unknown identity in a dataset to cell types, based on a provided list of marker genes,” says Elzbieta Gralinska of the Max Planck Institute for Molecular Genetics in Berlin.

The biotechnologist works in the team of Martin Vingron, which developed the technique, demonstrated its functionality on two publicly available datasets, and published the results in the Journal of Molecular Biology. Moreover, APL has been released as a free module for the statistical environment R. The APL package allows researchers to visually inspect their single-cell data and select individual genes with the cursor to learn more in-depth details.

Analyzing and grouping single cells

Why is it necessary to identify marker genes in the first place? Modern sequencing technologies are able to decipher individual RNA molecules in individual cells. From a blood sample, for example, each cell can be separated and a sample of the cell’s RNAs can be decoded. These single-cell data represent the active genes that were transcribed into RNA molecules.

The advantage: Instead of puzzling over which cell type a particular RNA belongs to, it can be traced back to its cell of origin. The disadvantage: sequencing thousands of RNAs in every single cell out of tens of thousands of cells produces extraordinary amounts of data.

One way out is to sort the cells based on their RNA content. “Single-cell data are composed of a wild mix of many different cell types. We are interested in cells of the same cell type, which should all behave similarly,” explains Martin Vingron. Hence, it makes sense to group similar cells computationally, he says. “For us, the marker genes define a cell type.”

Explore cell clusters interactively

Using publicly available data from white blood cells, the team demonstrated how the new algorithm works. The many different types of white blood cells like T-cells, B-cells, or monocytes are all grouped in separate clusters. The researchers confirmed known marker genes and were able to show that close relatives among the blood cells also share great similarity in their gene activity.

“Each of the marker genes we found with APL could have been discovered by at least one other existing method for identification of marker genes,” Gralinska says. But the advantage of APL over the existing algorithms is its graphical representation of the results, she says. “Existing tools provide long lists of genes and score values. Oftentimes, users go through the list and stop at an arbitrary cut-off,” Gralinska says.

In contrast, the new method provides a way to visualize these genes, click on each one and take a closer look at its activity, she says. “We're not just providing lists of marker genes, we're allowing users to review how these genes behave,” the researcher says. “With Association Plots, they can dive into their data to learn more about each cell type.” Plus, she says, it's very easy to break down the biological role of the most interesting genes in a subsequent step via Gene Ontology terms enrichment analysis, which is compatible with the APL software – something she considers “a very useful feature.”

The underlying mathematical model

The high-dimensional data that contain information on activity across genes cannot be represented visually without loss of information. The same is true for clustered data, all of which complicates analysis. “Our trick is that we take into account many more than just two or three dimensions, but ultimately create a two-dimensional diagram,” Gralinska says. 

The Association Plots are derived from a mathematical technique that simultaneously embeds both genes and cells in a common, high-dimensional space. Measuring the distances between genes and a given cell cluster in this space results in pairs of values that reflect the association of a gene to a given cluster and give insights into its association to other clusters.

“One shortcoming of APL is that we rely on pre-clustered data, which means we have to rely on other techniques for clustering,” says Martin Vingron. “Nevertheless, we hope that our new method will find many new users. We find that a visual and interactive process simply makes a better analysis.”

Max-Planck-Institut für molekulare Genetik

Recommend news PDF version / Print

Share on

Facts, background information, dossiers
  • genes
  • gene activity
  • data analysis
  • cell analysis
  • bioinformatics
More about MPI für molekulare Genetik
  • News

    More than the sum of mutations

    A new algorithm can predict which genes cause cancer, even if their DNA sequence is not changed. A team of researchers in Berlin combined a wide variety of data, analyzed it with “Artificial Intelligence” and identified numerous cancer genes. This opens up new perspectives for targeted canc ... more

    Breaks in the genome

    Breaks and rearrangements in the genome can lead to severe diseases, even if all genes remain intact. Hi-C, a method to map the three-dimensional structure of chromosomes, promises more reliable and accurate diagnoses of such defects, but is not used in the clinic yet. A team of researchers ... more

    Swiss Army Knife for Genome Research

    It is the the dream of every molecular geneticist: an easy-to-use program that compares data sets from different cellular conditions, identifies enhancer regions and then assigns them to their target genes. A research team led by Professor Martin Vingron at the Max Planck Institute for Mole ... more

More about Max-Planck-Gesellschaft
  • News

    More than microscopes can show

    A key protein for converting adult stem cells into cells that resemble embryonic stem cells has been visualized in unprecedented detail by an international team of researchers around Hans Schöler and Vlad Cojocaru of the Max Planck Institute for Molecular Biomedicine in Münster. By combing ... more

    Mirror image molecules reveal drought stress in forests

    Worldwide, plants emit about 100 million tonnes of monoterpenes into the atmosphere each year. These volatile organic molecules include many fragrances such as the molecule pinene – known for its pine fresh scent. Since these molecules are highly reactive and can form tiny aerosol particles ... more

    Pumping up the music of molecules

    Sensitive animal noses can sniff out trace particles, such as volatile organic compounds, in the ambient air. Humans, on the other hand, are developing innovative technologies for this purpose, such as optical spectroscopy. This uses laser light to detect the molecular composition of gases. ... more