Predicting sub-cellular localization of tRNA synthetases from their primary structures

Area of Science:

• Bioinformatics
• Molecular Biology
• Computational Biology

Background:

• Mitochondrial aminoacyl-tRNA synthetases (AARSs) are nucleus-encoded, synthesized in the cytosol, and imported into mitochondria.
• Distinguishing between cytosolic and mitochondrial AARSs is challenging due to similar physicochemical properties.
• Accurate sub-cellular localization prediction is vital for understanding mitochondrial protein synthesis.

Purpose of the Study:

• To develop and optimize computational methods for predicting the sub-cellular localization of AARSs.
• To identify distinguishable sequence patterns between cytosolic and mitochondrial AARS enzymes.
• To create a reliable tool for predicting AARS localization to aid in studying mitochondrial function.

Main Methods:

• Support Vector Machines (SVM) were employed for classification.
• Various sequence features were utilized, including amino acid composition, dipeptide composition, position-specific scoring matrices, and split amino acid composition (SAAC).
• Models were trained and validated using rigorous techniques like fivefold cross-validation on non-redundant datasets.

Main Results:

• SVM modules based on amino acid and dipeptide composition achieved high Mathews Correlation Coefficient (MCC) values (0.82 and 0.73).
• The split amino acid composition (SAAC) approach yielded significant improvements, reaching an MCC of 0.86.
• The optimized SA-SAAC method demonstrated superior performance, achieving an MCC of 0.92 (96.00% accuracy) during training and 0.95 (97.77% accuracy) on independent datasets.

Conclusions:

• The study successfully developed a highly accurate computational method (SA-SAAC) for predicting AARS sub-cellular localization.
• The developed web-server, MARSpred, provides a valuable resource for researchers studying mitochondrial protein synthesis and AARS function.
• Accurate prediction of AARS localization aids in understanding the complex interplay between nuclear and mitochondrial genomes in eukaryotic cells.