Identifying Drug Targets in Pancreatic Ductal Adenocarcinoma Through Machine Learning, Analyzing Biomolecular Networks, and Structural Modeling

Identifying Lymph Node Metastasis-Related Factors in Breast Cancer Using Differential Modular and Mutational Structural Analysis

Complex diseases are generally caused by disorders of biological networks and/or mutations in multiple genes. Comparisons of network topologies between different disease states can highlight key factors in their dynamic processes. Here, we propose a differential modular analysis approach that integrates protein-protein interactions with gene expression profiles for modular analysis, and introduces inter-modular edges and date hubs to identify the "core network module" that quantifies the significant phenotypic variation. Then, based on this core network module, key factors, including functional protein-protein interactions, pathways, and driver mutations, are predicted by the topological-functional connection score and structural modeling. We applied this approach to analyze the lymph node metastasis (LNM) process in breast cancer. The functional enrichment analysis showed that both inter-modular edges and date hubs play important roles in cancer metastasis and invasion, and in metastasis hallmarks. The structural mutation analysis suggested that the LNM of breast cancer may be the outcome of the dysfunction of rearranged during transfection (RET) proto-oncogene-related interactions and the non-canonical calcium signaling pathway via an allosteric mutation of RET. We believe that the proposed method can provide new insights into disease progression such as cancer metastasis.

MiR-200/183 family-mediated module biomarker for gastric cancer progression: an AI-assisted bioinformatics method with experimental functional survey

BACKGROUND

Gastric cancer (GC) is a major cancer burden throughout the world with a high mortality rate. The performance of current predictive and prognostic factors is still limited. Integrated analysis is required for accurate cancer progression predictive biomarker and prognostic biomarkers that help to guide therapy.

METHODS

An AI-assisted bioinformatics method that combines transcriptomic data and microRNA regulations were used to identify a key miRNA-mediated network module in GC progression. To reveal the module's function, we performed the gene expression analysis in 20 clinical samples by qRT-PCR, prognosis analysis by multi-variable Cox regression model, progression prediction by support vector machine, and in vitro studies to elaborate the roles in GC cells migration and invasion.

RESULTS

A robust microRNA regulated network module was identified to characterize GC progression, which consisted of seven miR-200/183 family members, five mRNAs and two long non-coding RNAs H19 and CLLU1. Their expression patterns and expression correlation patterns were consistent in public dataset and our cohort. Our findings suggest a two-fold biological potential of the module: GC patients with high-risk score exhibited a poor prognosis (p-value < 0.05) and the model achieved AUCs of 0.90 to predict GC progression in our cohort. In vitro cellular analyses shown that the module could influence the invasion and migration of GC cells.

CONCLUSIONS

Our strategy which combines AI-assisted bioinformatics method with experimental and clinical validation suggested that the miR-200/183 family-mediated network module as a "pluripotent module", which could be potential marker for GC progression.

HNRNPA2B1 as a potential therapeutic target for thymic epithelial tumor recurrence: An integrative network analysis

Thymic epithelial tumors (TETs) are rare malignant tumors, and the molecular mechanisms of both primary and recurrent TETs are poorly understood. Here we established comprehensive proteomic signatures of 15 tumors (5 recurrent and 10 non-recurrent) and 15 pair wised tumor adjacent normal tissues. We then proposed an integrative network approach for studying the proteomics data by constructing protein-protein interaction networks based on differentially expressed proteins and a machine learning-based score, followed by network modular analysis, functional enrichment annotation and shortest path inference analysis. Network modular analysis revealed that primary and recurrent TETs shared certain common molecular mechanisms, including a spliceosome module consisting of RNA splicing and RNA processing, but the recurrent TET was specifically related to the ribosome pathway. Applying the shortest path inference to the collected seed gene module identified that the ribonucleoprotein hnRNPA2B1 probably serves as a potential target for recurrent TET therapy. The drug repositioning combined molecular dynamics simulations suggested that the compound ergotamine could potentially act as a repurposing drug to treat recurrent TETs by targeting hnRNPA2B1. Our study demonstrates the value of integrative network analysis to understand proteotype robustness and its relationships with genotype, and provides hits for further research on cancer therapeutics.

Elevated ITGA2 expression promotes collagen type I-induced clonogenic growth of intrahepatic cholangiocarcinoma

Intrahepatic cholangiocarcinoma (iCCA) arises along the peripheral bile ducts and is often accompanied by a tumor microenvironment (TME) high in extracellular matrices (ECMs). In this study, we aimed to evaluate whether an ECM-rich TME favors iCCA progression. We identified ITGA2, which encodes collagen-binding integrin α2, to be differentially-expressed in iCCA tumors compared with adjacent normal tissues. Elevated ITGA2 is also positively-correlated with its ligand, collagen type I. Increased ITGA2 expression and its role in collagen type I binding was validated in vitro using four iCCA cell lines, compared with a non-cancerous, cholangiocyte cell line. Robust interaction of iCCA cells with collagen type I was abolished by either ITGA2 depletion or integrin α2β1-selective inhibitor treatment. In a phenotypic study, collagen type I significantly enhances clonogenic growth of HuCCA-1 and HuCCT-1 cells by three and sixfold, respectively. Inhibition of integrin α2 expression or its activity significantly blocks collagen type I-induced colony growth in both cell lines. Taken together, our data provide mechanistic evidence that collagen type I promotes growth of iCCA colonies through integrin α2 suggesting that the collagen type I-integrin α2 axis could be a promising target for cancer prevention and a therapeutic opportunity for this cancer.

Integrin subunit alpha V is a potent prognostic biomarker associated with immune infiltration in lower-grade glioma

As a member of integrin receptor family, ITGAV (integrin subunit α V) is involved in a variety of cell biological processes and overexpressed in various cancers, which may be a potential prognostic factor. However, its prognostic value and potential function in lower-grade glioma (LGG) are still unclear, and in terms of immune infiltration, it has not been fully elucidated. Here, the expression preference, prognostic value, and clinical traits of ITGAV were investigated using The Cancer Genome Atlas database (n = 528) and the Chinese Glioma Genome Atlas dataset (n = 458). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses and gene set enrichment analysis (GSEA) were used to explore the biological function of ITGAV. Using R package "ssGSEA" analysis, it was found thatthe ITGAV mRNA expression level showed intense correlation with tumor immunity, such as tumor-infiltrating immune cells and multiple immune-related genes. In addition, ITGAV is associated with some immune checkpoints and immune checkpoint blockade (ICB) and response to chemotherapy. and the expression of ITGAV protein in LGG patients was verified via immunohistochemistry (IHC). ITGAV expression was higher in LGG tissues than in normal tissues (P < 0.001) and multifactor analysis showed that ITGAV mRNA expression was an independent prognostic factor for LGG overall survival (OS; hazard ratio = 2.113, 95% confidence interval = 1.393-3.204, P < 0.001). GSEA showed that ITGAV expression was correlated with Inflammatory response, complement response, KRAS signal, and interferon response. ssGSEA results showed a positive correlation between ITGAV expression and Th2 cell infiltration level. ITGAV mRNA was overexpressed in LGG, and high ITGAV mRNA levels were found to be associated with poor protein expression and poor OS. ITGAV is therefore a potential biomarker for the diagnosis and prognosis of LGG and may be a potential immunotherapy target.

In Silico Methods for Identification of Potential Active Sites of Therapeutic Targets

Target identification is an important step in drug discovery, and computer-aided drug target identification methods are attracting more attention compared with traditional drug target identification methods, which are time-consuming and costly. Computer-aided drug target identification methods can greatly reduce the searching scope of experimental targets and associated costs by identifying the diseases-related targets and their binding sites and evaluating the druggability of the predicted active sites for clinical trials. In this review, we introduce the principles of computer-based active site identification methods, including the identification of binding sites and assessment of druggability. We provide some guidelines for selecting methods for the identification of binding sites and assessment of druggability. In addition, we list the databases and tools commonly used with these methods, present examples of individual and combined applications, and compare the methods and tools. Finally, we discuss the challenges and limitations of binding site identification and druggability assessment at the current stage and provide some recommendations and future perspectives.

lncRNAs Functioned as ceRNA to Sponge miR-15a-5p Affects the Prognosis of Pancreatic Adenocarcinoma and Correlates With Tumor Immune Infiltration

Pancreatic adenocarcinoma (PAAD) is one of the most common malignant tumors with poor prognosis worldwide. Mounting evidence suggests that the expression of lncRNAs and the infiltration of immune cells have prognostic value for patients with PAAD. We used Gene Expression Omnibus (GEO) database and identified six genes (COL1A2, ITGA2, ITGB6, LAMA3, LAMB3, and LAMC2) that could affect the survival rate of pancreatic adenocarcinoma patients. Based on a series of in silico analyses for reverse prediction of target genes associated with the prognosis of PAAD, a ceRNA network of mRNA (COL1A2, ITGA2, LAMA3, LAMB3, and LAMC2)–microRNA (miR-15a-5p)–long non-coding RNA (LINC00511, LINC01578, PVT1, and TNFRSF14-AS1) was constructed. We used the algorithm “CIBERSORT” to assess the proportion of immune cells and found three overall survival (OS)–associated immune cells (monocytes, M1 macrophages, and resting mast cell). Moreover, the OS-associated gene level was significantly positively associated with immune checkpoint expression and biomarkers of immune cells. In summary, our results clarified that ncRNA-mediated upregulation of OS-associated genes and tumor-infiltration immune cells (monocytes, M1 macrophages M1, and resting mast cell resting) correlated with poor prognosis in PAAD.

In silico Methods for Identification of Potential Therapeutic Targets

At the initial stage of drug discovery, identifying novel targets with maximal efficacy and minimal side effects can improve the success rate and portfolio value of drug discovery projects while simultaneously reducing cycle time and cost. However, harnessing the full potential of big data to narrow the range of plausible targets through existing computational methods remains a key issue in this field. This paper reviews two categories of in silico methods—comparative genomics and network-based methods—for finding potential therapeutic targets among cellular functions based on understanding their related biological processes. In addition to describing the principles, databases, software, and applications, we discuss some recent studies and prospects of the methods. While comparative genomics is mostly applied to infectious diseases, network-based methods can be applied to infectious and non-infectious diseases. Nonetheless, the methods often complement each other in their advantages and disadvantages. The information reported here guides toward improving the application of big data-driven computational methods for therapeutic target discovery.

Graphical abstract

The Highly Expressed IFIT1 in Nasopharyngeal Carcinoma Enhances Proliferation, Migration, and Invasion of Nasopharyngeal Carcinoma Cells

In this study, we aimed to identify potential targets modulating the progression of nasopharyngeal carcinoma (NPC) using integrated bioinformatics analysis and functional assays. Differentially expressed genes (DEGs) between NPC and normal tissues samples were obtained from publicly availably microarray datasets (GSE68799, GSE34573, and GSE53819) in the Gene Expression Omnibus (GEO) database. The bioinformatics analysis identified 49 common DEGs from three GEO datasets, which were mainly enriched in cytokine/chemokine pathways and extracellular matrix organization pathway. Further protein-protein interaction network analysis identified 11 hub genes from the 49 DEGs. The 11 hub genes were significantly up-regulated in the NPC tissues when compared to normal tissues by analyzing the Oncomine database. The 8 hub genes including COL5A1, COL7A1, COL22A1, CXCL11, IFI44L, IFIT1, RSAD2, and USP18 were significantly up-regulated in the NPC tissues when compared to normal tissues by using the Oncomine database. Further validation studies showed that IFIT1 was up-regulated in the NPC cells. Knockdown of IFI1T1 suppressed the proliferation, migration, and invasion of NPC cells; while IFIT1 overexpression promoted the proliferation, migration, and invasion of NPC cells. In conclusion, a total of 49 DEGs and 11 hub genes in NPC using the integrated bioinformatics analysis. IFIT1 was up-regulated in the NPC cells lines, and IFIT1 may act as an oncogene by promoting NPC cell proliferation, migration, and invasion.

Anticancer Effect of Heparin-Taurocholate Conjugate on Orthotopically Induced Exocrine and Endocrine Pancreatic Cancer

Simple Summary

Pancreatic cancer has a less than 9% 5-year survival rate among patients because it is very difficult to detect and diagnose early. Combinatorial chemotherapy with surgery or radiotherapy is a potential remedy to treat pancreatic cancer. However, these strategies still have side effects such as hair loss, skin soreness and fatigue. To overcome these side effects, angiogenesis inhibitors such as sunitinib are used to deliver targeted blood vessels around tumor tissues, including pancreatic cancer tumors. It is still controversial whether antiangiogenesis therapy is sufficient to treat pancreatic cancer. So far, many scientists have not been focused on the tumor types of pancreatic cancer when they have developed antipancreatic cancer medication. Here, we used heparin–taurocholate (LHT) as an anticancer drug to treat pancreatic cancer through inhibition of angiogenic growth factors. In this study, we examined the anticancer efficacy of LHT on various types of pancreatic cancer in an orthotopic model.

Abstract

Pancreatic cancers are classified based on where they occur, and are grouped into those derived from exocrine and those derived from neuroendocrine tumors, thereby experiencing different anticancer effects under medication. Therefore, it is necessary to develop anticancer drugs that can inhibit both types. To this end, we developed a heparin–taurocholate conjugate, i.e., LHT, to suppress tumor growth via its antiangiogenic activity. Here, we conducted a study to determine the anticancer efficacy of LHT on pancreatic ductal adenocarcinoma (PDAC) and pancreatic neuroendocrine tumor (PNET), in an orthotopic animal model. LHT reduced not only proliferation of cancer cells, but also attenuated the production of VEGF through ERK dephosphorylation. LHT effectively reduced the migration, invasion and tube formation of endothelial cells via dephosphorylation of VEGFR, ERK1/2, and FAK protein. Especially, these effects of LHT were much stronger on PNET (RINm cells) than PDAC (PANC1 and MIA PaCa-2 cells). Eventually, LHT reduced ~50% of the tumor weights and tumor volumes of all three cancer cells in the orthotopic model, via antiproliferation of cancer cells and antiangiogenesis of endothelial cells. Interestingly, LHT had a more dominant effect in the PNET-induced tumor model than in PDAC in vivo. Collectively, these findings demonstrated that LHT could be a potential antipancreatic cancer medication, regardless of pancreatic cancer types.

Centrality of drug targets in protein networks

Background

In the pharmaceutical industry, competing for few validated drug targets there is a drive to identify new ways of therapeutic intervention. Here, we attempted to define guidelines to evaluate a target’s ‘fitness’ based on its node characteristics within annotated protein functional networks to complement contingent therapeutic hypotheses.

Results

We observed that targets of approved, selective small molecule drugs exhibit high node centrality within protein networks relative to a broader set of investigational targets spanning various development stages. Targets of approved drugs also exhibit higher centrality than other proteins within their respective functional class. These findings expand on previous reports of drug targets’ network centrality by suggesting some centrality metrics such as low topological coefficient as inherent characteristics of a ‘good’ target, relative to other exploratory targets and regardless of its functional class. These centrality metrics could thus be indicators of an individual protein’s ‘fitness’ as potential drug target. Correlations between protein nodes’ network centrality and number of associated publications underscored the possibility of knowledge bias as an inherent limitation to such predictions.

Conclusions

Despite some entanglement with knowledge bias, like structure-oriented ‘druggability’ assessments of new protein targets, centrality metrics could assist early pharmaceutical discovery teams in evaluating potential targets with limited experimental proof of concept and help allocate resources for an effective drug discovery pipeline.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-021-04342-x.

Comprehensive analysis of abnormal expression, prognostic value and oncogenic role of the hub gene FN1 in pancreatic ductal adenocarcinoma via bioinformatic analysis and in vitro experiments

Background

Pancreatic ductal adenocarcinoma (PDAC) is one of the most commonly diagnosed cancers with a poor prognosis worldwide. Although the treatment of PDAC has made great progress in recent years, the therapeutic effects are still unsatisfactory. Methods. In this study, we identified differentially expressed genes (DEGs) between PDAC and normal pancreatic tissues based on four Gene Expression Omnibus (GEO) datasets (GSE15471, GSE16515, GSE28735 and GSE71729). A protein–protein interaction (PPI) network was established to evaluate the relationship between the DEGs and to screen hub genes. The expression levels of the hub genes were further validated through the Gene Expression Profiling Interactive Analysis (GEPIA), ONCOMINE and Human Protein Atlas (HPA) databases, as well as the validation GEO dataset GSE62452. Additionally, the prognostic values of the hub genes were evaluated by Kaplan–Meier plotter and the validation GEO dataset GSE62452. Finally, the mechanistic roles of the most remarkable hub genes in PDAC were examined through in vitro experiments.

Results

We identified the following nine hub genes by performing an integrated bioinformatics analysis: COL1A1, COL1A2, FN1, ITGA2, KRT19, LCN2, MMP9, MUC1 and VCAN. All of the hub genes were significantly upregulated in PDAC tissues compared with normal pancreatic tissues. Two hub genes (FN1 and ITGA2) were associated with poor overall survival (OS) rates in PDAC patients. Finally, in vitro experiments indicated that FN1 plays vital roles in PDAC cell proliferation, colony formation, apoptosis and the cell cycle.

Conclusions

In summary, we identified two hub genes that are associated with the expression and prognosis of PDAC. The oncogenic role of FN1 in PDAC was first illustrated by performing an integrated bioinformatic analysis and in vitro experiments. Our results provide a fundamental contribution for further research aimed finding novel therapeutic targets for overcoming PDAC.

Identification of hub genes and their novel diagnostic and prognostic significance in pancreatic adenocarcinoma

OBJECTIVE

The main reasons for the poor prognoses of pancreatic adenocarcinoma (PA) patients are rapid early-stage progression, advanced stage metastasis, and chemotherapy resistance. Identification of novel diagnostic and prognostic biomarkers of PA is therefore urgently needed.

METHODS

Three mRNA microarray datasets were obtained from the Gene Expression Omnibus database to select differentially expressed genes (DEGs). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses for hub genes were performed using DAVID. Correlations between expression levels of hub genes and cancer-infiltrating immune cells were investigated by TIMER. Cox proportional hazard regression analyses were also performed. Serum hub genes were screened using the HPA platform and verified for diagnostic value using ELISAs.

RESULTS

We identified 59 hub genes among 752 DEGs. GO analysis indicated that these 59 hub genes were mainly involved in the defense response to viruses and the type I interferon signaling pathway. We also discovered that RSAD2 and SMC4 were associated with immune cell infiltration in the PA microenvironment. Additionally, DLGAP5 mRNA might be used as an independent risk factor for the prognoses of PA patients. Furthermore, the protein encoded by ISG15, which exists in peripheral blood, was validated as a potential diagnostic biomarker that distinguished PA patients from healthy controls (area under the curve: 0.902, 95% confidence interval: 0.819-0.961).

CONCLUSIONS

Our study suggested that RSAD2 and SMC4 were associated with immune cell infiltration in the PA microenvironment, while DLGAP5 mRNA expression might be an independent risk factor for the survival prognoses of PA patients. Moreover, ELISAs indicated that serum ISG15 could be a potential novel diagnostic biomarker for PA.

A quantitative score of immune cell infiltration predicts the prognosis in pancreatic ductal adenocarcinoma

Pancreatic Ductal Adenocarcinoma (PDAC) is characterized by an extensive and dense fibrous stroma, which plays an active role in tumor growth and metastasis. Despite the growing importance of the tumor microenvironment in PDAC prognosis, the immune cell infiltration landscape of PDAC has not been elucidated. In this study, we applied a credible computational algorithm to comprehensively estimate the immune cell infiltration (ICI) patterns of 876 PDAC patients. Two ICI phenotypes were identified, and a ICIscore was constructed using ssGSEA algorithm. The ICIscore could significantly predict the prognosis and chemotherapy benefit of PDAC patients in both the discovery and the five validation cohorts. Multivariate cox analysis also identified the independent predictive role of the ICIscore in PDAC prognosis. A high ICIscore subtype was characterized by immune-active signaling pathways and anti-tumor immunity while a low ICIscore subtype was associated with tumor progressive signaling pathways. Four immunotherapy cohorts further supported the use of the ICIscore as a prognostic biomarker for patients receiving immune checkpoint inhibitors in other cancer types. The ICIscore reveals a close relationship between the ICI environment and prognosis and may provide new treatment strategies for PDAC patients.

Identification of the Neurokinin-1 Receptor as Targetable Stratification Factor for Drug Repurposing in Pancreatic Cancer

The SP/NK1R-complex plays an important role in tumor proliferation. Targeting of the neurokinin-1 receptor in previous studies with its antagonist aprepitant (AP) resulted in anti-tumoral effects in colorectal cancer and hepatoblastoma. However, there is still a lack of knowledge regarding its effects on pancreatic cancer. Therefore, we treated human pancreatic ductal adenocarcinoma (PDAC) cell lines (Capan-1, DanG, HuP-T3, Panc-1, and MIA PaCa-2) and their cancer stem cell-like cells (CSCs) with AP and analyzed functional effects by MTT-, colony, and sphere formation assays, respectively; moreover, we monitored downstream mechanisms by flow cytometry. NK1R inhibition resulted in dose-dependent growth reduction in both CSCs and non-CSCs without induction of apoptosis in most PDAC cell lines. More importantly, we identified striking AP dependent cell cycle arrest in all parental cells. Furthermore, gene expression and the importance of key genes in PDAC tumorigenesis were analyzed combining RT-qPCR in eight PDAC cell lines with publicly available datasets (TCGA, GEO, CCLE). Surprisingly, we found a better overall survival in patients with high NK1R levels, while at the same time, NK1R was significantly decreased in PDAC tissue compared to normal tissue. Interestingly, there is currently no differentiation between the isoforms of NK1R (truncated and full; NK1R-tr and -fl) in any of the indicated public transcriptomic records, although many publications already emphasize on important regulatory differences between the two isoforms of NK1R in many cancer entities. In conclusion, analysis of splice variants might potentially lead to a stratification of PDAC patients for NK1R-directed therapies. Furthermore, we presume PDAC patients with high expressions of NK1R-tr might benefit from treatment with AP to improve chemoresistance. Therefore, analysis of splice variants might potentially lead to a stratification of PDAC patients for NK1R-directed therapies.

Network-Based Target Prioritization and Drug Candidate Identification for Multiple Sclerosis: From Analyzing "Omics Data" to Druggability Simulations

Multiple sclerosis (MS) is the most common chronic inflammatory demyelinating disease of the central nervous system. While the drugs currently available for MS provide symptomatic benefit, there is no curative treatment. The emergence of large-scale multiomics data and network theory provide new opportunities for drug discovery in MS, as these are promising strategies for developing novel drugs. In this study, we proposed a computational framework that combined biomolecular network modeling and structural dynamics analysis to facilitate the discovery of new drugs with potential activity in MS. First, we developed a new shortest path-based algorithm that prioritized differentially expressed genes using a newly topological and functional exploration of protein-protein interaction network. Then, pathway enrichment analysis and an assessment of target druggability suggested that TNF-α-induced protein 3 (TNFAIP3), which is involved in NF-κ B signaling, could be a potential therapeutic target for MS. Finally, druggability simulations and mutation enrichment analysis of the TNFAIP3 dimer presented two druggable sites. Follow-up pharmacophore model-based virtual screening of the two sites yielded 30 hit compounds with low energy scores. In summary, this novel method based on analyzing "omics data" and performing druggability simulations, is a systematic approach that unravels disease mechanisms and links them to the chemical space to develop treatments and can be applied to other complex diseases.

Knowledge-Guided "Community Network" Analysis Reveals the Functional Modules and Candidate Targets in Non-Small-Cell Lung Cancer

Non-small-cell lung cancer (NSCLC) represents a heterogeneous group of malignancies that are the leading cause of cancer-related death worldwide. Although many NSCLC-related genes and pathways have been identified, there remains an urgent need to mechanistically understand how these genes and pathways drive NSCLC. Here, we propose a knowledge-guided and network-based integration method, called the node and edge Prioritization-based Community Analysis, to identify functional modules and their candidate targets in NSCLC. The protein–protein interaction network was prioritized by performing a random walk with restart algorithm based on NSCLC seed genes and the integrating edge weights, and then a “community network” was constructed by combining Girvan–Newman and Label Propagation algorithms. This systems biology analysis revealed that the CCNB1-mediated network in the largest community provides a modular biomarker, the second community serves as a drug regulatory module, and the two are connected by some contextual signaling motifs. Moreover, integrating structural information into the signaling network suggested novel protein–protein interactions with therapeutic significance, such as interactions between GNG11 and CXCR2, CXCL3, and PPBP. This study provides new mechanistic insights into the landscape of cellular functions in the context of modular networks and will help in developing therapeutic targets for NSCLC.