Model Running

Job id:

   Cut-off:

Strand:

2. Paste in a query sequence to find the R-loop formation propensit:y

- Multiple sequences (less than 5) are allowed if separated by lines starting
with '>' followed by the sequence name.
- Prediction results will only be kept for 30 days.

3. File upload : Rather than pasting a query sequence, you can choose to upload a text file containing the sequence(s).

4. Run it locally : You may get our code from github to run it locally.
Model Introduction

DeepER is a deep learning-based tool to predict R-loop forming sequences. The basic framework of DeepER includes one layer of Bi-LSTM and two layers of Bi-LSTM with residual blocks, followed by a fully connected layer activated by sigmoid. Base-level probability of R-loop fromation will be predicted for a given 5kb-long sequence. A sliding window approach (window size = 200bp and step size = 10bp) is then applied to find R-loop-forming regions, defined as consecutive sequences showing average probability >= 0.95 (default cut_off).

© 2020-2025 NANJING UNIVERSITY. ALL RIGHTS RESERVED.
CONDITIONS OF USE