Cells differentiated from embryonic stem cells are generally phenotypically immature, so identifying what cell type they are can be a bit like trying to figure out what exactly the picture your child drew for you in kindergarten is supposed to be. “That’s lovely darling but…” Luckily, for biologists at least, a team of Dutch researchers led by Susana Chuva de Sousa Lopes have come up with a solution.
KeyGenes is an algorithm that can predict the identity of a test tissue from its transcriptional profile based on next generation sequencing (NGS) data of 21 human fetal and extra-embryonic tissues from the first and second trimester of development. First, Lopes’ team used a training set of 76 tissues to identify so called ‘classifier’ genes. These genes were highly expressed throughout development in the tissues they characterize and were either not expressed or only lowly expressed in several other tissues, which helped to define a sort of transcriptional “barcode” for each tissue. Then, the classifier genes were used to predict the identity of samples in a test set.
- KeyGenes performed extremely well in the first test set, correctly identifying 38/39 fetal tissues.
- More rigorous testing based on published microarray and NGS datasets showed that the developmental classifier genes could accurately predict the identity of their adult organ counterparts.
- When applied to cells in culture, KeyGenes correctly predicted the differentiation of human pluripotent cells to a specific cell type. Moreover, the identity ‘score’ depended on the number of passages in culture, showing that KeyGenes can be used to optimize the quality of differentiated stem cells.
This is not the first algorithm to predict stem cell fate based on transcriptional profiling, but as Lopes points out, previous algorithms were based on the analysis of adult tissues with microarrays, which often excludes many classifier genes and are less relevant to early development.
Besides optimizing protocols for stem cell differentiation, KeyGenes may also help to researchers to understand how and why development goes wrong in some cases by identifying genes that are expressed in the wrong place or at the wrong time.
Check out the divining powers of KeyGenes at Stem Cell Reports, May 2015.