Protein Complexes Identification with Family-Wise Error Rate Control.

Area of Science:

• Bioinformatics
• Systems Biology
• Computational Biology

Background:

• Protein complex detection is crucial in bioinformatics and systems biology.
• Existing methods lack strict control over the family-wise error rate (FWER) for protein complex identification.
• There is a need for methods that provide a more rigorous statistical foundation for detecting significant protein complexes.

Purpose of the Study:

• To propose a new method, SSF, for detecting statistically significant protein complexes.
• To ensure strict control of the family-wise error rate (FWER) for each identified protein complex.
• To enhance the accuracy and reliability of protein complex detection in biological networks.

Main Methods:

• Developed a p-value calculation method using Fisher's exact test to assess protein-complex associations.
• Implemented a seed expansion procedure for searching significant protein complexes.
• Utilized a set cover strategy for effective redundancy elimination within detected complexes.

Main Results:

• The SSF method achieved the highest precision across five benchmark datasets.
• SSF outperformed three competing methods in Normalized Mutual Information (NMI) and F1 score.
• Experimental results demonstrate the superior performance and statistical rigor of the proposed SSF algorithm.

Conclusions:

• The SSF method offers a statistically sound approach to protein complex detection with FWER control.
• SSF provides improved accuracy and performance compared to existing state-of-the-art methods.
• This work advances the field of protein complex identification by introducing a more reliable and precise detection strategy.