Carboxylator: incorporating solvent-accessible surface area for identifying protein carboxylation sites

摘要: In proteins, glutamate (Glu) residues are transformed into γ-carboxyglutamate (Gla) in a process called carboxylation. The of protein carboxylation catalyzed by γ-glutamyl carboxylase is deemed to be important due its involvement biological processes such as blood clotting cascade and bone growth. There an increasing interest within the scientific community identify sites. However, experimental identification sites via mass spectrometry-based methods observed expensive, time-consuming, labor-intensive. Thus, we were motivated design computational method for identifying This work aims investigate considering composition amino acids that surround modification With implication modified residue prefers accessible on surface protein, solvent-accessible area (ASA) around also investigated. Radial basis function network then employed build predictive model using various features Based five-fold cross-validation evaluation, trained combined acid sequence (AA20D), composition, ASA, yields highest accuracy at 0.874. Furthermore, independent test done involving data not included indicates silico feasible means preliminary analysis. Additionally, presented this implemented Carboxylator (http://csb.cse.yzu.edu.tw/Carboxylator/), web-based tool carboxylated proteins with order help users investigating