A sequence-based prediction of Kruppel-like factors proteins using XGBoost and optimized features

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Krüppel-like factors (KLF) are crucial transcription factors regulating cell processes.
• Existing bioinformatics methods aid KLF protein research.
• Novel computational approaches are needed to enhance KLF identification.

Purpose of the Study:

• To develop a novel computational method for identifying KLF proteins.
• To utilize machine learning on primary sequence features for KLF prediction.
• To provide a tool for biologists and researchers studying KLF proteins.

Main Methods:

• Feature calculation from primary protein sequences.
• Development of an XGBoost-based machine learning model.
• Model validation using an independent dataset.

Main Results:

• The XGBoost model achieved 96.4% accuracy and 0.704 MCC in identifying KLF proteins.
• The model demonstrated promising performance on an independent dataset.
• The developed computational approach is efficient for KLF identification.

Conclusions:

• The proposed machine learning model is an effective tool for identifying KLF proteins.
• This approach can assist researchers in discovering new KLF proteins.
• The study provides valuable computational resources for KLF research.