Weidong Tian

Lung Adenocarcinoma From East Asian Never-Smokers Is a Disease Largely Defined by Targetable Oncogenic Mutant Kinases

PURPOSE To determine the proportion of lung adenocarcinomas from East Asian never-smokers who harbor known oncogenic driver mutations. PATIENTS AND METHODS In this surgical series, 52 resected lung adenocarcinomas from never-smokers (< 100 cigarettes in a lifetime) at a single institution (Fudan University, Shanghai, China) were analyzed concurrently for mutations in EGFR, KRAS, NRAS, HRAS, HER2, BRAF, ALK, PIK3CA, TP53 and LKB1. RESULTS Forty-one tumors harbored EGFR mutations, three harbored EML4-ALK fusions, two harbored HER2 insertions, and one harbored a KRAS mutation. All mutations were mutually exclusive. Thus, 90% (47 of 52; 95% CI, 0.7896 to 0.9625) of lung adenocarcinomas from never-smokers were found to harbor well-known oncogenic mutations in just four genes. No BRAF, NRAS, HRAS, or LKB1 mutations were detected, while 15 had TP53 mutations. Four tumors contained PIK3CA mutations, always together with EGFR mutations. CONCLUSION To our knowledge, this study represents the first comprehensive and concurrent analysis of major recurrent oncogenic mutations found in a large cohort of lung adenocarcinomas from East Asian never-smokers. Since drugs are now available that target mutant EGFR, HER2, and ALK, respectively, this result indicates that prospective mutation testing in these patients should successfully assign a targeted therapy in the majority of cases.

A critical assessment of Mus musculus gene function prediction using integrated genomic evidence

Background:Several years after sequencing the human genome and the mouse genome, much remains to be discovered about the functions of most human and mouse genes. Computational prediction of gene function promises to help focus limited experimental resources on the most likely hypotheses. Several algorithms using diverse genomic data have been applied to this task in model organisms; however, the performance of such approaches in mammals has not yet been evaluated.Results:In this study, a standardized collection of mouse functional genomic data was assembled; nine bioinformatics teams used this data set to independently train classifiers and generate predictions of function, as defined by Gene Ontology (GO) terms, for 21,603 mouse genes; and the best performing submissions were combined in a single set of predictions. We identified strengths and weaknesses of current functional genomic data sets and compared the performance of function prediction algorithms. This analysis inferred functions for 76% of mouse genes, including 5,000 currently uncharacterized genes. At a recall rate of 20%, a unified set of predictions averaged 41% precision, with 26% of GO terms achieving a precision better than 90%.Conclusion:We performed a systematic evaluation of diverse, independently developed computational approaches for predicting gene function from heterogeneous data sources in mammals. The results show that currently available data for mammals allows predictions with both breadth and accuracy. Importantly, many highly novel predictions emerge for the 38% of mouse genes that remain uncharacterized.

Epigenetic features are significantly associated with alternative splicing

BackgroundWhile alternative splicing (AS) contributes greatly to protein diversities, the relationship between various types of AS and epigenetic factors remains largely unknown.ResultsIn this study, we discover that a number of epigenetic features, including DNA methylation, nucleosome occupancy, specific histone modifications and protein features, are strongly associated with AS. To further enhance our understanding of the association between these features and AS, we cluster our investigated features based on their association patterns with each AS type into four groups, with H3K36me3, EGR1, GABP, SRF, SIN3A and RNA Pol II grouped together and showing strongest association with AS. In addition, we find that the AS types can be classified into two general classes, namely the exon skipping related process (ESRP), and the alternative splice site selection process (ASSP), based on their association levels with the epigenetic features.ConclusionOur analysis thus suggests that epigenetic features are likely to play important roles in regulating AS.

A network-based gene-weighting approach for pathway analysis

Classical algorithms aiming at identifying biological pathways significantly related to studying conditions frequently reduced pathways to gene sets, with an obvious ignorance of the constitutive non-equivalence of various genes within a defined pathway. We here designed a network-based method to determine such non-equivalence in terms of gene weights. The gene weights determined are biologically consistent and robust to network perturbations. By integrating the gene weights into the classical gene set analysis, with a subsequent correction for the “over-counting” bias associated with multi-subunit proteins, we have developed a novel gene-weighed pathway analysis approach, as implemented in an R package called “Gene Associaqtion Network-based Pathway Analysis” (GANPA). Through analysis of several microarray datasets, including the p53 dataset, asthma dataset and three breast cancer datasets, we demonstrated that our approach is biologically reliable and reproducible, and therefore helpful for microarray data interpretation and hypothesis generation.

High precision multi-genome scale reannotation of enzyme function by EFICAz

BackgroundThe functional annotation of most genes in newly sequenced genomes is inferred from similarity to previously characterized sequences, an annotation strategy that often leads to erroneous assignments. We have performed a reannotation of 245 genomes using an updated version of EFICAz, a highly precise method for enzyme function prediction.ResultsBased on our three-field EC number predictions, we have obtained lower-bound estimates for the average enzyme content in Archaea (29%), Bacteria (30%) and Eukarya (18%). Most annotations added in KEGG from 2005 to 2006 agree with EFICAz predictions made in 2005. The coverage of EFICAz predictions is significantly higher than that of KEGG, especially for eukaryotes. Thousands of our novel predictions correspond to hypothetical proteins. We have identified a subset of 64 hypothetical proteins with low sequence identity to EFICAz training enzymes, whose biochemical functions have been recently characterized and find that in 96% (84%) of the cases we correctly identified their three-field (four-field) EC numbers. For two of the 64 hypothetical proteins: PA1167 from Pseudomonas aeruginosa, an alginate lyase (EC 4.2.2.3) and Rv1700 of Mycobacterium tuberculosis H37Rv, an ADP-ribose diphosphatase (EC 3.6.1.13), we have detected annotation lag of more than two years in databases. Two examples are presented where EFICAz predictions act as hypothesis generators for understanding the functional roles of hypothetical proteins: FLJ11151, a human protein overexpressed in cancer that EFICAz identifies as an endopolyphosphatase (EC 3.6.1.10), and MW0119, a protein of Staphylococcus aureus strain MW2 that we propose as candidate virulence factor based on its EFICAz predicted activity, sphingomyelin phosphodiesterase (EC 3.1.4.12).ConclusionOur results suggest that we have generated enzyme function annotations of high precision and recall. These predictions can be mined and correlated with other information sources to generate biologically significant hypotheses and can be useful for comparative genome analysis and automated metabolic pathway reconstruction.

An iterative network partition algorithm for accurate identification of dense network modules

A key step in network analysis is to partition a complex network into dense modules. Currently, modularity is one of the most popular benefit functions used to partition network modules. However, recent studies suggested that it has an inherent limitation in detecting dense network modules. In this study, we observed that despite the limitation, modularity has the advantage of preserving the primary network structure of the undetected modules. Thus, we have developed a simple iterative Network Partition (iNP) algorithm to partition a network. The iNP algorithm provides a general framework in which any modularity-based algorithm can be implemented in the network partition step. Here, we tested iNP with three modularity-based algorithms: multi-step greedy (MSG), spectral clustering and Qcut. Compared with the original three methods, iNP achieved a significant improvement in the quality of network partition in a benchmark study with simulated networks, identified more modules with significantly better enrichment of functionally related genes in both yeast protein complex network and breast cancer gene co-expression network, and discovered more cancer-specific modules in the cancer gene co-expression network. As such, iNP should have a broad application as a general method to assist in the analysis of biological networks.

FuncBase: a resource for quantitative gene function annotation

Summary: Computational gene function prediction can serve to focus experimental resources on high-priority experimental tasks. FuncBase is a web resource for viewing quantitative machine learning-based gene function annotations. Quantitative annotations of genes, including fungal and mammalian genes, with Gene Ontology terms are accompanied by a community feedback system. Evidence underlying function annotations is shown. For example, a custom Cytoscape viewer shows functional linkage graphs relevant to the gene or function of interest. FuncBase provides links to external resources, and may be accessed directly or via links from species-specific databases. Availability: FuncBase as well as all underlying data and annotations are freely available via http://func.med.harvard.edu/ Contact: fritz_roth@hms.harvard.edu

Combining Hi-C data with phylogenetic correlation to predict the target genes of distal regulatory elements in human genome

Defining the target genes of distal regulatory elements (DREs), such as enhancer, repressors and insulators, is a challenging task. The recently developed Hi-C technology is designed to capture chromosome conformation structure by high-throughput sequencing, and can be potentially used to determine the target genes of DREs. However, Hi-C data are noisy, making it difficult to directly use Hi-C data to identify DRE–target gene relationships. In this study, we show that DREs–gene pairs that are confirmed by Hi-C data are strongly phylogenetic correlated, and have thus developed a method that combines Hi-C read counts with phylogenetic correlation to predict long-range DRE–target gene relationships. Analysis of predicted DRE–target gene pairs shows that genes regulated by large number of DREs tend to have essential functions, and genes regulated by the same DREs tend to be functionally related and co-expressed. In addition, we show with a couple of examples that the predicted target genes of DREs can help explain the causal roles of disease-associated single-nucleotide polymorphisms located in the DREs. As such, these predictions will be of importance not only for our understanding of the function of DREs but also for elucidating the causal roles of disease-associated noncoding single-nucleotide polymorphisms.

Integration of cancer gene co-expression network and metabolic network to uncover potential cancer drug targets.

Cell metabolism is critical for cancer cell transformation and progression. In this study, we have developed a novel method, named Met-express, that integrates a cancer gene co-expression network with the metabolic network to predict key enzyme-coding genes and metabolites in cancer cell metabolism. Met-express successfully identified a group of key enzyme-coding genes and metabolites in lung, leukemia, and breast cancers. Literature reviews suggest that approximately 33-53% of the predicted genes are either known or suggested anti-cancer drug targets, while 22% of the predicted metabolites are known or high-potential drug compounds in therapeutic use. Furthermore, experimental validations prove that 90% of the selected genes and 70% of metabolites demonstrate the significant anti-cancer phenotypes in cancer cells, implying that they may play important roles in cancer metabolism. Therefore, Met-express is a powerful tool for uncovering novel therapeutic biomarkers.

Explaining the disease phenotype of intergenic SNP through predicted long range regulation.

Thousands of disease-associated SNPs (daSNPs) are located in intergenic regions (IGR), making it difficult to understand their association with disease phenotypes. Recent analysis found that non-coding daSNPs were frequently located in or approximate to regulatory elements, inspiring us to try to explain the disease phenotypes of IGR daSNPs through nearby regulatory sequences. Hence, after locating the nearest distal regulatory element (DRE) to a given IGR daSNP, we applied a computational method named INTREPID to predict the target genes regulated by the DRE, and then investigated their functional relevance to the IGR daSNPs disease phenotypes. 36.8% of all IGR daSNP-disease phenotype associations investigated were possibly explainable through the predicted target genes, which were enriched with, were functionally relevant to, or consisted of the corresponding disease genes. This proportion could be further increased to 60.5% if the LD SNPs of daSNPs were also considered. Furthermore, the predicted SNP-target gene pairs were enriched with known eQTL/mQTL SNP-gene relationships. Overall, its likely that IGR daSNPs may contribute to disease phenotypes by interfering with the regulatory function of their nearby DREs and causing abnormal expression of disease genes.