New technique improves efficiency and accuracy of single-cell RNA sequencing

In the era of personalized medicine, scientists are using new genetic and genomic knowledge to help them determine the best treatment for a given patient. In the case of cancer, the first step towards these treatments is a survey of the behavior of tumor cells to determine the best drugs to use to attack them.

The researchers then use sequencing of DNA and RNA to examine cell populations and determine which genes are expressed in a sample of cancerous tissue. However, traditional sequencing methods may obscure the fact that not all tumor cells behave in the same way. Not recognizing this means that if you target a tumor with a specific type of drug, some cells may be different enough to survive and thrive.

As part of a major breakthrough in genomics, it is now possible to examine what a cell does at a given moment with a technique called single-cell RNA sequencing (cRNA-seq). This method examines the amount of messenger RNA (mRNA) in a cell and compares these to other cells in order to look for differences in gene expression.

However, the information you find may depend on how you run your test and how the data is analyzed. Lana Garmire, Ph.D., associate professor in the Department of Computer Medicine and Bioinformatics at Michigan Medicine and her team are studying ways to eliminate some biases that may make it difficult to interpret ARNs data. -seq.

"A lot of the noise in this type of sequencing comes from the fact that you have to measure samples in extremely low quantities and in different batches," she says. For example, the tissue sample analyzed by the researcher may not fit onto a plate, a piece of equipment used to store cell samples, and must therefore be divided into two plates. The differences resulting from this division are called batch effects. Genomics researchers need to correct these batch effects, but this process can be a problem: how do you know if a difference is a batch effect or a real difference between cells?

New uses for data

Bioinformatics is the term used to collect and analyze complex biological data using computer programs. It is a relatively new field, born of the ability to collect huge amounts of biological data, such as DNA sequences and proteins.

Researchers rely on bioinformatic techniques to determine which genes are expressed in single cells. But they had to work around the noise introduced by different research protocols and batch effects. Garmire, who recently joined the UM from the University of Hawaii and is the new director of the Faculty of Bioinformatics of the University of Michigan's Faculty of Medicine, discovered a way to more efficient in identifying differences between cells using the same set of data produced during sequencing experiments. Instead of relying on gene expression, she discovered that examining so-called single-nucleotide variants (SNVs) can eliminate some of this uncertainty. "With SNV, you are dealing with binary numbers, 0 and 1. Either the mutation is there or not."

Recall that the genes consist of nucleotides represented by the letters A, T, G and C which constitute a code which is translated into a protein. The Garmire method looks for differences between single nucleotides, knowing that an A can only be replaced by a T and a G by a C. This new work, described in Nature Communications, has developed a new set of procedures for processing scRNA-seq data and retrieving this variant information. In addition, by using a computer program called SSrGE, they can link this variant information to more traditional gene expression information.

"This gives us information about different subpopulations of tumor cells and becomes a sort of fingerprint that can be tagged to identify cell-to-cell differences," says Garmire.

What does it all mean

In the end, drug manufacturers and clinicians use these targets to guide pharmaceutical treatments. "When you want to tackle the problem, you are tackling its fundamental characteristics: mutations – clinicians may be able to use this information later to guide their treatment." Garmire is eager to bring bioinformatics into the lab, to help researchers who accumulate large amounts of data to use them and develop downstream clinical applications. "We divide the body and specialize, but at the end of the day, you have to look holistically and ask what I do and who does it help to do?" We develop computer tools to get researchers into bioinformatics , bench scientists and clinicians connect the dots and finally make the change.