Predicting Human Transcription Factor Interactions
Author Information
Author(s): Schmeier Sebastian, Jankovic Boris, Bajic Vladimir B.
Primary Institution: Computational Bioscience Research Center (CBRC), King Abdullah University of Science and Technology (KAUST)
Hypothesis
Can we predict interactions between human transcription factors using their primary structure?
Conclusion
The proposed method achieves a prediction accuracy of 85.39% for human transcription factor interactions.
Supporting Evidence
- The model achieved an accuracy of 85.39% on a blind set of interactions.
- Feature selection identified 97 model features that improved prediction performance.
- The method simplifies the prediction process by using only primary sequence information.
Takeaway
The study created a way to guess how human proteins called transcription factors interact with each other, and it works pretty well.
Methodology
The method uses primary sequence information of transcription factors and employs quadratic discriminant analysis for prediction.
Potential Biases
Potential bias due to the limited number of known interactions and the method of generating negative examples.
Limitations
The study is limited by the small number of known transcription factor interactions and the challenge of creating a reliable negative dataset.
Participant Demographics
The study focuses on human transcription factors.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website