DeepSELEX

Amir Engel

Amir Engel

Unknown

0 0
  • 0 Collaborators

The binding of DNA by proteins, known as transcription factors (TF), is a central mechanism in gene regulation. DeepSELEX is an advanced algorithm to infer intrinsic DNA-binding preferences using deep neural networks. ...learn more

Project status: Under Development

Artificial Intelligence

Intel Technologies
DevCloud

Overview / Usage

Transcription factor DNA-binding is a central mechanism in gene regulation. Biologists would

like to know where and when these factors bind DNA. Hence, they require accurate DNA-binding models

to enable binding prediction to any DNA sequence. Recent technological advancements can measure the

binding of a single transcription factor to thousands of DNA sequences. One of the prevailing techniques is

high-throughput SELEX, which measures protein-DNA binding by high-throughput sequencing over several

cycles of enrichment. Unfortunately, current computational methods to infer the binding preferences from

high-throughput SELEX data do not exploit the richness of these data, and are under-using the most

advanced computational technique.

Methodology / Approach

Objectives: Improvement of the DeepSELEX algorithm - by improving the aggregation function of the algorithm, using KL divergence for choosing the most suitable enrichment cycle, implementing deep learning tools that haven’t been used in the architecture of the network such as dropout and batch normalization.

The innovation: Improving the aggregation function of the algorithm, choosing the most suitable enrichment cycle by using KL divergence. Using the second enrichment cycle of the HT-SELEX in the aggregation function which does not use it in the current architecture (or determine it is unnecessary). Using deep learning tools for regulation which haven’t been used in the architecture of the network such as dropout and batch normalization. Using more data that derives from various experiments, or decide that the extra data is unnecessary, which could help in the experiment.

Technologies Used

Deep Neural network , CNN

Comments (0)