transcription factor technology


Mogrify technology was developed as a systematic means of identifying the optimal combination of transcription factors required to convert any cell type into any other cell type. The figure below shows the step-wise scientific approach used in order to achieve such predictions. Firstly, the gene expression levels are compared in the source and target cell types (originally using the FANTOM5 dataset, now using advanced sources of genetic and epigenetic input data). This step determines the changes in the gene expression levels that are needed in order to achieve the conversion. Secondly, transcription factors are scored compared to the required changes in gene expression levels by both direct and indirect influences (using the MARA and STRING databases). This includes the regulatory network information on protein-DNA interactions as well as protein-protein interactions. Thirdly, local transcriptomic regulatory networks are built to calculate the effect of the transcription factors on the gene expression levels. Lastly, the optimal combination of transcription factors is chosen in order to achieve a minimum of 98% of the necessary changes in gene expression (identified in step 1) for the cell conversion to occur.

In addition to identifying transcription factor driven cell conversions, small molecules that are known to affect the expression of the key predicted transcription factors can be identified from published literature to create a small molecule conversion cocktail. This has the added benefit of not requiring the transduction of the transcription factors and consequently holds greater potential as an in vivo reprograming therapy.

transcription factor
(a) Differential gene expression levels are identified by comparing RNA sequencing data, while using the FANTOM5 database as background.
transcription factor
(b) All human transcription factors are ranked according to their effect on the identified differential gene expression using both direct (DNA-protein interactions, MARA database) and indirect effects (protein-protein interactions, STRING database).
transcription factor
(c) The optimal combination of transcription factors is identified in order to maximize the differential gene expression while minimizing the overlap in effects.

Images from Rackham OJL et al. A predictive computational framework for direct reprograming between human cell types. Nature Genetics (2016)