Discovery of Hidden Gene Regulators: A Novel Machine Learning Approach to Transcriptional Pause Site Determination
ISEF · 2019 Computational Biology and Bioinformatics
Overview
Transcriptional pausing, the inhibition of RNA polymerase (RNAP) elongation during transcription, is a vital regulator of DNA transcription that offers increased control over gene expression. Erroneous or absent pausing can lead to genetic diseases such as Down Syndrome, Cystic Fibrosis, Huntington’s Disease, and Sickle Cell Anemia. Currently, the only way to determine pause sites in a genome is to spend multiple days and thousands of dollars on genome sequencing, which still produces extremely inaccurate results. The objective of this study was to develop a more effective and less expensive way to determine pause sites. Using previously supported pause sites in the E. coli genome and aggregating sequential and parametric DNA data, this study developed and optimized probabilistic and neural machine learning models to predict pause site incidence in DNA segments. A randomized trees algorithm and novel sequential memory model proved to accurately predict pause site locations at an unprecedented 0.995 confidence level across all test environments. In addition, the pause site predictions of these models were then back-tested through in vivo biochemical experiments that analyzed the effect of inserting the predicted pause sequences on the expression of Green Fluorescent Protein (GFP) in a non-pathogenic strain of E. Coli, resulting in a strong correlation between real and predicted locations. The results of this study ultimately provide a highly favorable alternative to pause site prediction. Furthermore, gaining insight into the relationship between transcriptional activities and the human body can potentially help discover the genes that are directly related to genetic diseases.
Awards (1)
- Second Award of $1,500 $1,500
Competition history
- ISEF 2019
Resources
Related projects
ISEF · 2022
RiboBayes: Assessing the Transcriptome-Wide Expression of Ribosome Pause Sites in Ribosome Profiling Data With Bayesian Wavelet Thresholding
ISEF · 2019
Deep Learning to Evaluate the Combinatorial Impact of Genetic Variants on Gene Expression
ISEF · 2018
Illuminating Gene Dysregulation in Cancer: Deep Learning Identification of Disrupted Transcription Factor Binding Sites
ISEF · 2014
RNNScan: Eukaryotic Gene Finding via Hybrid Recurrent Neural Networks
Closest projects by meaning, across every fair and year in the corpus.
Source: Regeneron International Science and Engineering Fair