Human cells ‘fingerprint’: scPred Method Identifies Type of Different Cells
A new method scPred analyzes data from individual human cells. This method could be game-changing for diagnosing some of the most devastating diseases including autoimmune disease and cancer.
At the Garvan Institute of Medical Research, a research team has developed a method to ‘fingerprint’ human cells by the combination of machine learning algorithms and single-cell analysis.
The new method discovered has the potential to detect cancer earlier, help personalize treatments to individual patients and can identify the cells at the root of autoimmune disease. This method is called ‘scPred’.
The lead of the study, Joseph Powell, the Associate Professor Director at the Garvan-Weizmann Centre for Cellular Genomics says, ” In medical diagnostics, we are at the beginning of a significant new frontier by developing this new way to identify very specific types of cells.
Having a closer look at human cells
Associate Professor Powell explains, ” Based on a limited number of markers found on the cell surface or inside the cell, we have classified the main different cells in the human body for a long time now. But now, we are seeing a variety of cell types underneath just one ‘type’, for example, only a subgroup of cancer cells may actually form a metastatic tumor even though different cancer cells could all have the same cell surface markers.”
Extensive information on what makes a cell unique was discovered by the new method of analyzing transcripts of individual cells—a measure of which genes are active in different cells.
Within the huge number of transcripted data’s information, the challenge of determining what can provide the most useful information that defines a cell type is solved by the scPred method.
Rather than estimating 20,000 things at once, this method looks at which patterns of those 20,000 have the most predictive power in distinguishing one cell type from another cell type. All the transcript data from a single cell is first collected in this scPred method.
A Ph.D. student at the University of Queensland, José Alquicira-Hernández, the first author the study explains, “To test what features make a specific cell type the most distinct from another cell, a statistical model on those patterns is done by scPred, this is almost like a unique fingerprint of each cell.”
A new dimension on diagnostics
Scientists can use the trained model once a certain cell type has been ‘fingerprinted’ to look for the same type of cell in datasets or any other samples from anywhere in the world. Collaborators at Stanford University in the United States have analyzed datasets of colorectal cancer cells using the scPred approach and validated this method. With over 98% accuracy, the researchers were able to identify cancer cells from a tissue sample using the scPred models.
The scientists say that using their new method, a lot of improvements in the resolution of cell types can be done and it can also uncover diseased cells that are outside the scope of current medical diagnostics.
Translation to patients
For the first time, this new method opens the technology to diagnostic applications by allowing researchers to take snapshots of over 20,000 different pieces of information in a single cell’s transcript, thanks to advanced single-cell sequencing methods.
Scientists are moving to the next phase of translating the method to accredited tests for clinical practice through the Garvan-Weizmann Centre for Cellular Genomics.
Associate Professor Powell says, ” The possibility of earlier detection is possible by our scPred method, it could let us determine the stage of a cancer patient, whether their tumor cells have signatures that indicate resistance to chemotherapy or to what potential drugs they will respond to.
The journal Genome Biology published this research study.
Editors Note- A New Method to Human cells ‘fingerprint’: scPred, Human cells ‘fingerprint’: scPred Method.
Author: Prathibha HC