Loop Prediction

CLAMP: Predicting Specific Protein-Mediated Chromatin Loops in Diverse Species with a Language Model of Chromatin Accessibility

About This Tool

This tool allows you to predict specific protein-mediated chromatin loops in diverse species using our language model of chromatin accessibility. You can input DNA region pairs and our model will predict the likelihood of loop formation between these regions.

Due to limited online server resources, CLAMP⁺ supports up to 10 pairs per submission. If you need to predict more loops, please deploy CLAMP model on your server/cluster.

According to our test, the anlysis time is about 10 seconds per pair.

Select Experiment Parameters:

These options mean selecting a CLAMP model that we have pre-trained ahead of time on a species, markers or cell lines

Species

Please select a species

Marker

Please select a marker

Cell Line

Please select a cell line

Type

Please select a type

Please select parameters in order: Species → Marker → Cell Line → Type. Options are filtered based on available data.

Input candidate anchor pairs:

Or upload a file in csv format:

Supported file format: CSV (.csv)

Context-specific epigenetic data (Optional):

The epigenetic data you upload should correspond to the same species and preferably a similar cell line as the selected CLAMP model for optimal performance.

Supported file formats: signal p-value BigWig (.bw, .bigWig). Max file size: 512MB.
Note: Epigenetic sequencing data should be processed using the official ENCODE pipeline (https://www.encodeproject.org/atac-seq/)

Email Address (Optional)

We will notify you once your job is completed.

Example Result

Need Help?

If you encounter any issues or have questions about using this tool, please refer to our documentation or contact our support team.