BITFAM infers transcription factor activity in individual cells

Monika R, PSG College of Technology, Coimbatore

Transcriptional gene regulation is how a cell regulates the conversion of DNA to RNA, and it is also a complex process that utilizes a network of interactions. This process is primarily controlled by DNA binding proteins known as transcription factors (TFs)/regulators, that attach to specific DNA sequences and thereby repress or initiate gene expression (i.e.) it regulates genes to be turned ‘on’ and ‘off’ to make sure that they’re expressed within the right cell at an accurate time and amount throughout the lifetime of the cell and the organism.

Transcriptional regulatory networks control the expression levels of thousands of genes as a part of diverse biological processes. To understand the complex regulatory mechanisms in cellular systems, inferring the relationships between TFs and their targets is very important. With the rapid advancement of bioscience, tremendous amounts of biological data are being produced by high-throughput technologies. 

How can we determine transcription factor activity (TFA)?

Most of the techniques made an implicit assumption that ‘the expression of the TF-gene can be used as a proxy for TFA’. But this assumption was wrong because various post-transcriptional and post-translational modifications abolish significant correlations between TFAs and the level of TF-gene expression. Owing to post-translational modifications, TFAs are difficult to measure experimentally and so, many analyses infer TFAs by computational analysis.

To overcome the problem of the dearth of correlation, several research groups have proposed statistical approaches that infer TFAs. Recently proposed approaches treat the true TFAs as latent variables that are inferred from observed expression data of their gene targets. These techniques require prior knowledge of the network topology. However, in the majority of cases, some topology has been elucidated via techniques like gene knockouts and ChIP. Chromatin immunoprecipitation (ChIP) assay is one important technique that identifies TFs and their target genes in several organisms.

Inferring the TFA in individual cells would allow researchers to review activity profiles in all the major cell types of organs like the heart, brain, or lungs. Bayesian Inference Transcription Factor Activity Model (BITFAM), identifies meaningful TFAs in single cells and also provides valuable insights into underlying TF regulatory mechanisms.

How does BITFAM work? 

BITFAM is a software tool, which helps researchers to detect gene regulations and is developed by a team of scientists at the University of Illinois Chicago. This system works by decomposing single-cell transcriptomic profile data gathered from scRNA-seq into TFAs, with existing biological data on TF-target genes gathered from ChIP-seq. 

The system also predicts algorithmically, which TFs are presumably to be more active in individual cells, by ranking TF target genes for each scRNA-seq data set. By predicting the more active one among every TF in a cell, it can provide researchers a good idea like which TF to look at first when exploring new drug targets to work on that cell type. 

BITFAM also performs downstream analyses, like clustering of cell subpopulations using inferred TFAs. Most of the techniques/models to infer TFAs fail to deal with at least one among the subsequent objectives: 

• Understand the fact that TF is subjected to post-transcriptional regulation;
• TFs cooperate as a functional complex in gene expression regulation;
• Provide a learning algorithm and a model with manageable computational complexity. 

But BITFAM not only infers TFAs but also addresses these three issues. BITFAM can learn target genes of the TF specific to their biological function and would also allow for the inference of preferred target genes and functions of TF.

BITFAM’s use in future

In computational biology, statistical methods for inferring TFAs are an important area of research. This can be due to their ability to gather the information that isn’t readily available through standard experimental practice. Believing that the time has arrived for these methods to become standard software utilized in biological laboratories to enhance experimental work, much in the way that sequence alignment tools are now routinely employed by experimentalists. By providing a simple yet powerful implementation of an already tried and tested method, we can hope that BITFAM will become accessible and useful to a large community of scientists engaged in gene regulation.

Also read: Recent insights into the role of Ubiquitination

Sources:

• Gao, S., Dai, Y., & Rehman, J. (2021). A Bayesian inference transcription factor activity model for the analysis of single-cell transcriptomes. Genome Research, 31(7), 1296–1311. https://doi.org/10.1101/gr.265595.120
• Chicago, U. of I. at. (n.d.-a). Machine learning algorithm predicts how genes are regulated in individual cells. Retrieved July 2, 2021, from https://phys.org/news/2021-06-machine-algorithm-genes-individual-cells.html

• The Corrosion Prediction from the Corrosion Product Performance
• Nitrogen Resilience in Waterlogged Soybean plants
• Cell Senescence in Type II Diabetes: Therapeutic Potential
• Transgene-Free Canker-Resistant Citrus sinensis with Cas12/RNP
• AI Literacy in Early Childhood Education: Challenges and Opportunities

About the author: Monika R is an enthusiastic Biotech student aspiring for an opportunity to develop skills and grow professionally in the research field. Extremely motivated and possess strong interpersonal skills and the ability to understand concepts quickly.