In This Section

How Can Single-cell Sequencing Overcome Computational Limitations?

Published on January 30, 2023 in Cornerstone Blog · Last updated 6 months ago
AddtoAny
Share:

WATCH THIS PAGE

Subscribe to be notified of changes or updates to this page.

1 + 5 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.
A newly developed single-cell multimodal deep clustering software uses machine learning to analyze data about a single cells multiple characteristics.

A newly developed single-cell multimodal deep clustering software uses machine learning to analyze data about a single cell’s multiple characteristics.

The findings

Researchers from Children's Hospital of Philadelphia and New Jersey Institute of Technology (NJIT) developed new software that integrates multiple data sources from a single cell. This method provides important clues about how one change can lead to several others, enabling researchers to pinpoint the causes of genetic-based diseases.

Why it matters

With single-cell sequencing technologies, researchers can study gene expression, as well as messenger RNA, proteins, and even organelles in great detail. This is particularly relevant in cancer research, since the information can be used to determine the disease's genetic evolution. However, combining different cell characteristics for clustering analysis — for example, how a genetic variant might directly impact messenger RNA, protein synthesis, or epigenetics — presents a computational challenge. The CHOP/NJIT team's single-cell multimodal deep clustering (scMDC) software uses machine learning to analyze data about a single cell's multiple characteristics. 

Who conducted the study

Hakon Hakonarson, MD, PhD, director of the Center for Applied Genomics, was first author, and Tian Tian, PhD, a postdoctoral researcher in the Center, co-authored the study. 

How they did it

To test the clustering performance of the new model, the researchers employed simulation and real-data experiments to compare it against competing methods. The results showed that scMDC outperformed existing single cell single-modal and multimodal clustering approaches. The new software also utilizes linear scalability, meaning that more data sources provided to the scMDC yield better results.

Quick thoughts

"With this tool, we can better understand a single cell as an entity and not just as a fragmented unit," Dr. Hakonarson said.

What's next

"This is a significant advancement and allows us to integrate and put all of this information into biological perspective, which is particularly important when considering information on different diseases," Dr. Hakonarson said.

Where the study was published

The study appeared in Nature Communications.

Where to learn more

Read more about the Center for Applied Genomics.