RFQAmodel

RFQAmodel (Random Forest Quality Assessment) is a software tool that assesses the quality of protein structure models generated by template-free prediction methods. The tool aims to help users select the best model from a large set of generated models and determine whether the chosen model is likely correct.

Key features of the model:

1. Combines existing quality assessment scores with two predicted contact map alignment scores to identify correct models that other methods may miss.

2. Trained on diverse protein domains, balanced for features that affect modeling success.

3. Tested on a separate set of 244 protein domains with similar properties.

4. Rank models according to the model score, assigning high-confidence scores to correct models and identifying incorrect models.

5. Outperforms other methods in identifying correct models when only a few of the generated models are accurate.

6. Demonstrates similar performance on CASP12 and CASP13 free-modeling targets.

7. Can be used iteratively to focus computational efforts on challenging modeling targets by generating models until a high-confidence correct model is produced.