MetalTrans: A Biological Language Model-Based Approach for Predicting Disease-Associated Mutations in Protein Metal-Binding Sites.

The critical importance of accurately predicting mutations in protein metal-binding sites for advancing drug discovery and enhancing disease diagnostic processes cannot be overstated. In response to this imperative, MetalTrans emerges as an accurate predictor for disease-associated mutations in protein metal-binding sites. The core innovation of MetalTrans lies in its seamless integration of multifeature splicing with the Transformer framework, a strategy that ensures exhaustive feature extraction. Central to MetalTrans's effectiveness is its deep feature combination strategy, which merges evolutionary-scale modeling amino acid embeddings with ProtTrans embeddings, thus shedding light on the biochemical properties of proteins. Employing the Transformer component, MetalTrans leverages the self-attention mechanism to delve into higher-level representations. Utilizing mutation site information for feature fusion not only enriches the feature set but also sidesteps the common pitfall of overestimation linked to protein sequence-based predictions. This nuanced approach to feature fusion is a key differentiator, enabling MetalTrans to outperform existing methods significantly, as evidenced by comparative analyses. Our evaluations across varied metal binding site data sets (specifically Zn, Ca, Mg, and Mix) underscore MetalTrans's superior performance, which achieved the average AUC values of 0.971, 0.965, 0.980, and 0.945 on multiple 5-fold cross-validation, respectively. Remarkably, against the multichannel convolutional neural network method on a benchmark independent test set, MetalTrans demonstrated unparalleled robustness and superiority, boasting the AUC score of 0.998 on multiple 5-fold cross-validation. Our comprehensive examination of the predicted outcomes further confirms the effectiveness of the model. The source codes, data sets, and prediction results for MetalTrans can be accessed for academic usage at https://github.com/EduardWang/MetalTrans.

Role of K364 next to the active site cysteine in voltage-dependent phosphatase activity of Ci-VSP.

Voltage-sensing phosphatase (VSP) consists of the voltage sensor domain (VSD) similar to that of voltage-gated ion channels and the cytoplasmic phosphatase region with remarkable similarity to the phosphatase and tensin homolog deleted on chromosome 10 (PTEN). Membrane depolarization activates VSD, leading to dephosphorylation of three species of phosphoinositides (phosphatidylinositol phosphates (PIPs)), PI(3,4,5)P3, PI(4,5)P2, and PI(3,4)P2. VSP dephosphorylates 3- and 5-phosphate of PIPs, unlike PTEN, which shows rigid 3-phosphate specificity. In this study, a bioinformatics search showed that some mammals have VSP orthologs with amino acid diversity in the active center motif, Cx5R, which is highly conserved among protein tyrosine phosphatases and PTEN-related phosphatases; lysine next to the active site cysteine in the Cx5R motif was substituted for methionine in VSP orthologs of Tasmanian devil, koala, and prairie deer mouse, and leucine in opossum. Since lysine at the corresponding site in PTEN is known to be critical for enzyme activities, we attempted to address the significance of amino acid diversity among VSP orthologs at this site. K364 was changed to different amino acids in sea squirt VSP (Ci-VSP), and voltage-dependent phosphatase activity in Xenopus oocyte was studied using fluorescent probes for PI(4,5)P2 and PI(3,4)P2. All mutants retained both 5-phosphatase and 3-phosphatase activity, indicating that lysine at this site is dispensable for 3-phosphatase activity, unlike PTEN. Notably, K364M mutant showed increased activity both of 5-phosphatase and 3-phosphatase compared with the wild type (WT). It also showed slower kinetics of voltage sensor motion. Malachite green assay of K364M mutant did not show significant difference of phosphatase activity from WT, suggesting tighter interaction between substrate binding and voltage sensing. Mutation corresponding to K364M in the zebrafish VSP led to enhanced voltage-dependent dephosphorylation of PI(4,5)P2. Further studies will provide clues to understanding of substrate preference in PIPs phosphatases as well as to customization of a molecular tool.