ENTPRISE-X: Predicting disease-associated frameshift and nonsense mutations

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Next Generation Sequencing generates vast data, necessitating accurate identification of disease-causing genetic mutations.
• Genome-wide association studies are limited to common variations and cannot identify personal or rare disease-associated mutations.
• The significance of identified genetic variations often remains unknown, hindering clinical application.

Purpose of the Study:

• To extend the machine learning approach ENTPRISE for predicting the disease association of missense mutations to include frameshift and nonsense mutations.
• To develop a novel computational tool, ENTPRISE-X, for more accurate identification of disease-causing genetic variations.
• To provide a freely available web server for academic users to facilitate research.

Main Methods:

• Extension of the machine learning-based ENTPRISE approach to accommodate frameshift and nonsense mutations.
• Comparative performance analysis against state-of-the-art methods VEST-indel and DDIG-in.
• Large-scale testing on the ExAC dataset to evaluate prediction accuracy and false positive rates.

Main Results:

• ENTPRISE-X significantly outperforms VEST-indel and DDIG-in in predicting the disease association of germline frameshift mutations, achieving a Matthew's correlation coefficient (MCC) of 0.586.
• ENTPRISE-X demonstrates a substantially lower fraction of predicted disease-causing variations (16%) compared to VEST-indel (26%) and DDIG-in (65%) on the ExAC dataset.
• The developed method shows improved accuracy and reduced false positives in classifying genetic variations.

Conclusions:

• ENTPRISE-X represents a significant advancement in computational prediction of disease-associated genetic mutations, particularly for frameshift and nonsense types.
• The tool offers improved accuracy and specificity compared to existing methods, aiding in the discovery of personal disease drivers.
• The availability of the ENTPRISE-X web server will accelerate research in precision medicine and genetic disease diagnostics.