This form allows you to predict the quality of 3D models for a given protein target. Further information and references can be found on the ModFOLD home page. Before you submit a prediction please refer to the help page. Click 'Help' in each section for detailed instructions.

Required - Input sequence of protein target (single letter code) Help

Required - Upload model/models (either a single PDB file or a tarred and gzipped directory of PDB files) Help


Select global accuracy score optimisation preference Help
Ranking (for multiple models, the most accurate models are ranked at the top)
Balanced (default, good correlations and model ranking)
Correlation (predicted global scores correlate well with the observed scores)

Optional - E-mail address (you will be sent a link to graphical and machine readable results when the job is completed) Help

Optional - Short name for protein target Help

Required - "I understand that any personal data collected from me in order to provide this service will be used by the University of Reading for no other purpose and will be deleted as soon as it is no longer needed." Help