Electron transfer coupling is a critical factor in determining electron transfer rates. This coupling strength can be sensitive to details in molecular geometries, especially intermolecular configurations. Thus, studying charge transporting behavior with a full first-principle approach demands a large amount of computation resources in quantum chemistry (QC) calculation. To address this issue, we developed a machine learning (ML) approach to evaluate electronic coupling. A prototypical ML model for an ethylene system was built by kernel ridge regression with Coulomb matrix representation. Since the performance of the ML models highly dependent on their building strategies, we systematically investigated the generality of the ML models, the choice of features and target labels. The best ML model trained with 40 000 samples achieved a mean absolute error of 3.5 meV and greater than 98% accuracy in predicting phases. The distance and orientation dependence of electronic coupling was successfully captured. Bypassing QC calculation, the ML model saved 10−10 4 times the computation cost. With the help of ML, reliable charge transport models and mechanisms can be further developed.
The process of nucleocytoplasmic shuttling is mediated by karyopherins. Dysregulated expression of karyopherins may trigger oncogenesis through aberrant distribution of cargo proteins. Karyopherin subunit alpha-2 (KPNA2) was previously identified as a potential biomarker for nonsmall cell lung cancer by integration of the cancer cell secretome and tissue transcriptome data sets. Knockdown of KPNA2 suppressed the proliferation and migration abilities of lung cancer cells. However, the precise molecular mechanisms underlying KPNA2 activity in cancer remain to be established. In the current study, we applied gene knockdown, subcellular fractionation, and stable isotope labeling by amino acids in cell culturebased quantitative proteomic strategies to systematically analyze the KPNA2-regulating protein profiles in an adenocarcinoma cell line. Interaction network analysis revealed that several KPNA2-regulating proteins are involved in the cell cycle, DNA metabolic process, cellular component movements and cell migration. Importantly, E2F1 was identified as a potential novel cargo of KPNA2 in the nuclear proteome. The mRNA levels of potential effectors of E2F1 measured using quantitative PCR indicated that E2F1 is one of the "master molecule" responses to KPNA2 knockdown. Immunofluorescence staining and immunoprecipitation assays disclosed colocalization and association between E2F1 and KPNA2. An in vitro protein binding assay further demonstrated that E2F1 interacts directly with KPNA2. Moreover, knockdown of KPNA2 led to subcellular redistribution of E2F1 in lung cancer cells. Our results collectively demonstrate the utility of quantitative proteomic approaches and provide a fundamental platform to further explore the biological roles of KPNA2 in nonsmall cell lung cancer. Molecular & Cellular
Quantum chemistry calculations have been very useful in providing many key detailed properties and enhancing our understanding of molecular systems. However, such calculation, especially with ab initio models, can be time-consuming. For example, in the prediction of charge-transfer properties, it is often necessary to work with an ensemble of different thermally populated structures. A possible alternative to such calculations is to use a machine-learning based approach. In this work, we show that the general prediction of electronic coupling, a property that is very sensitive to intermolecular degrees of freedom, can be obtained with artificial neural networks, with improved performance as compared to the popular kernel ridge regression method. We propose strategies for optimizing the learning rate and batch size, improving model performance, and further evaluating models to ensure that the physical signatures of charge-transfer coupling are well reproduced. We also address the effect of feature representation as well as statistical insights obtained from the loss function and the data structure. Our results pave the way for designing a general strategy for training such neural-network models for accurate prediction.
Nonsmall cell lung cancer (NSCLC) is the most common type of lung cancer, which is one of the most prominent causes of cancer-related mortality worldwide. Discovery of serum tumor markers could facilitate early NSCLC detection and metastatic prognosis. Here, we simultaneously analyzed the NSCLC cell secretome and proteomic profiles of pleural effusion from lung adenocarcinoma patients for NSCLC biomarker discovery. Retinoblastoma-associated binding protein 46 (RbAp46), one of the proteins detected both in NSCLC cell secretome and pleural effusion proteome, was chosen for further evaluation. Both of RbAp46 mRNA and protein levels were upregulated significantly in NSCLC cancer tissues. Serum levels of RbAp46 were markedly higher in NSCLC patients than in healthy controls, and a combination of RbAp46 and CEA could outperform CEA alone in discriminating NSCLC patients from healthy persons. Importantly, elevated serum RbAp46 level was highly correlated with NSCLC distant metastasis. Moreover, knockdown of RbAp46 inhibited the migration ability of lung cancer cells. Our data collectively suggest that RbAp46 serves as a novel biomarker and prognosticator for NSCLC, and is involved in lung cancer cell migration.
Malignant pleural effusion (MPE) obtained from lung adenocarcinoma may contain potentially useful biomarkers for detection of lung cancer. In this study, we used a removal system for high-abundance proteins followed by one-dimensional SDS-PAGE combined with nano-LC-MS/MS to generate a comprehensive MPE proteome data set with 482 nonredundant proteins. Next, we integrated the MPE proteome and secretome data sets from three adenocarcinoma cell lines, with a view to identifying potential PE biomarkers originating from malignant cells. Four potential candidates, alpha-2-HS-glycoprotein (AHSG), angiogenin, cystatin-C, and insulin-like growth factor-binding protein 2, (IGFBP2), were isolated for preclinical validation using ELISA. Both AHSG and IGFBP2 levels were increased in lung patients with MPE (n = 68), compared to those with nonmalignant pleural effusion (n = 119). Notably, the IGFBP2 level was higher in MPE, compared with that in benign diseases (bacteria pneumonia and tuberculosis pleuritis), and significantly associated with malignancy, regardless of the cancer type. Our data additionally support an extracellular function of IGFBP2 in migration in lung cancer cells. These findings collectively suggest that the adenocarcinoma MPE proteome provides a useful data set for malignancy biomarker research.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.