A computational method to aid the design and analysis of single cell RNA-seq experiments for cell type identification.

BACKGROUND: The advent of single cell RNA sequencing (scRNA-seq) enabled researchers to study transcriptomic activity within individual cells and identify inherent cell types in the sample. Although numerous computational tools have been developed to analyze single cell transcriptomes, there are no published studies and analytical packages available to guide experimental design and to devise suitable analysis procedure for cell type identification. RESULTS: We have developed an empirical methodology to address this important gap in single cell experimental design and analysis into an easy-to-use tool called SCEED (Single Cell Empirical Experimental Design and analysis). With SCEED, user can choose a variety of combinations of tools for analysis, conduct performance analysis of analytical procedures and choose the best procedure, and estimate sample size (number of cells to be profiled) required for a given analytical procedure at varying levels of cell type rarity and other experimental parameters. Using SCEED, we examined 3 single cell algorithms using 48 simulated single cell datasets that were generated for varying number of cell types and their proportions, number of genes expressed per cell, number of marker genes and their fold change, and number of single cells successfully profiled in the experiment. CONCLUSIONS: Based on our study, we found that when marker genes are expressed at fold change of 4 or more, either Seurat or SIMLR algorithm can be used to analyze single cell dataset for any number of single cells isolated (minimum 1000 single cells were tested). However, when marker genes are expected to be only up to fold change of 2, choice of the single cell algorithm is dependent on the number of single cells isolated and rarity of cell types to be identified. In conclusion, our work allows the assessment of various single cell methods and also aids in the design of single cell experiments.

MultiDCoX: Multi-factor analysis of differential co-expression.

BACKGROUND: Differential co-expression (DCX) signifies change in degree of co-expression of a set of genes among different biological conditions. It has been used to identify differential co-expression networks or interactomes. Many algorithms have been developed for single-factor differential co-expression analysis and applied in a variety of studies. However, in many studies, the samples are characterized by multiple factors such as genetic markers, clinical variables and treatments. No algorithm or methodology is available for multi-factor analysis of differential co-expression. RESULTS: We developed a novel formulation and a computationally efficient greedy search algorithm called MultiDCoX to perform multi-factor differential co-expression analysis. Simulated data analysis demonstrates that the algorithm can effectively elicit differentially co-expressed (DCX) gene sets and quantify the influence of each factor on co-expression. MultiDCoX analysis of a breast cancer dataset identified interesting biologically meaningful differentially co-expressed (DCX) gene sets along with genetic and clinical factors that influenced the respective differential co-expression. CONCLUSIONS: MultiDCoX is a space and time efficient procedure to identify differentially co-expressed gene sets and successfully identify influence of individual factors on differential co-expression.

Systems consequences of amplicon formation in human breast cancer.

Chromosomal structural variations play an important role in determining the transcriptional landscape of human breast cancers. To assess the nature of these structural variations, we analyzed eight breast tumor samples with a focus on regions of gene amplification using mate-pair sequencing of long-insert genomic DNA with matched transcriptome profiling. We found that tandem duplications appear to be early events in tumor evolution, especially in the genesis of amplicons. In a detailed reconstruction of events on chromosome 17, we found large unpaired inversions and deletions connect a tandemly duplicated ERBB2 with neighboring 17q21.3 amplicons while simultaneously deleting the intervening BRCA1 tumor suppressor locus. This series of events appeared to be unusually common when examined in larger genomic data sets of breast cancers albeit using approaches with lesser resolution. Using siRNAs in breast cancer cell lines, we showed that the 17q21.3 amplicon harbored a significant number of weak oncogenes that appeared consistently coamplified in primary tumors. Down-regulation of BRCA1 expression augmented the cell proliferation in ERBB2-transfected human normal mammary epithelial cells. Coamplification of other functionally tested oncogenic elements in other breast tumors examined, such as RIPK2 and MYC on chromosome 8, also parallel these findings. Our analyses suggest that structural variations efficiently orchestrate the gain and loss of cancer gene cassettes that engage many oncogenic pathways simultaneously and that such oncogenic cassettes are favored during the evolution of a cancer.

An oestrogen-receptor-alpha-bound human chromatin interactome.

Genomes are organized into high-level three-dimensional structures, and DNA elements separated by long genomic distances can in principle interact functionally. Many transcription factors bind to regulatory DNA elements distant from gene promoters. Although distal binding sites have been shown to regulate transcription by long-range chromatin interactions at a few loci, chromatin interactions and their impact on transcription regulation have not been investigated in a genome-wide manner. Here we describe the development of a new strategy, chromatin interaction analysis by paired-end tag sequencing (ChIA-PET) for the de novo detection of global chromatin interactions, with which we have comprehensively mapped the chromatin interaction network bound by oestrogen receptor alpha (ER-alpha) in the human genome. We found that most high-confidence remote ER-alpha-binding sites are anchored at gene promoters through long-range chromatin interactions, suggesting that ER-alpha functions by extensive chromatin looping to bring genes together for coordinated transcriptional regulation. We propose that chromatin interactions constitute a primary mechanism for regulating transcription in mammalian genomes.

Adaptive Sentinel Testing in Workplace for COVID-19 Pandemic.

Testing and isolation of infectious employees is one of the critical strategies to make the workplace safe during the pandemic for many organizations. Adaptive testing frequency reduces cost while keeping the pandemic under control at the workplace. However, most models aimed at estimating test frequencies were structured for municipalities or large organizations such as university campuses of highly mobile individuals. By contrast, the workplace exhibits distinct characteristics: employee positivity rate may be different from the local community because of rigorous protective measures at workplace, or self-selection of co-workers with common behavioral tendencies for adherence to pandemic mitigation guidelines. Moreover, dual exposure to COVID-19 occurs at work and home that complicates transmission modeling, as does transmission tracing at the workplace. Hence, we developed bi-modal SEIR (Susceptible, Exposed, Infectious, and Removed) model and R-shiny tool that accounts for these differentiating factors to adaptively estimate the testing frequency for workplace. Our tool uses easily measurable parameters: community incidence rate, risks of acquiring infection from community and workplace, workforce size, and sensitivity of testing. Our model is best suited for moderate-sized organizations with low internal transmission rates, no-outward facing employees whose position demands frequent in-person interactions with the public, and low to medium population positivity rates. Simulations revealed that employee behavior in adherence to protective measures at work and in their community, and the onsite workforce size have large effects on testing frequency. Reducing workplace transmission rate through workplace mitigation protocols and higher sensitivity of the test deployed, although to a lesser extent. Furthermore, our simulations showed that sentinel testing leads to only marginal increase in the number of infections even for high community incidence rates, suggesting that this may be a cost-effective approach in future pandemics. We used our model to accurately guide testing regimen for three campuses of the Jackson Laboratory.

A Genomically and Clinically Annotated Patient-Derived Xenograft Resource for Preclinical Research in Non-Small Cell Lung Cancer.

Patient-derived xenograft (PDX) models are an effective preclinical in vivo platform for testing the efficacy of novel drugs and drug combinations for cancer therapeutics. Here we describe a repository of 79 genomically and clinically annotated lung cancer PDXs available from The Jackson Laboratory that have been extensively characterized for histopathologic features, mutational profiles, gene expression, and copy-number aberrations. Most of the PDXs are models of non-small cell lung cancer (NSCLC), including 37 lung adenocarcinoma (LUAD) and 33 lung squamous cell carcinoma (LUSC) models. Other lung cancer models in the repository include four small cell carcinomas, two large cell neuroendocrine carcinomas, two adenosquamous carcinomas, and one pleomorphic carcinoma. Models with both de novo and acquired resistance to targeted therapies with tyrosine kinase inhibitors are available in the collection. The genomic profiles of the LUAD and LUSC PDX models are consistent with those observed in patient tumors from The Cancer Genome Atlas and previously characterized gene expression-based molecular subtypes. Clinically relevant mutations identified in the original patient tumors were confirmed in engrafted PDX tumors. Treatment studies performed in a subset of the models recapitulated the responses expected on the basis of the observed genomic profiles. These models therefore serve as a valuable preclinical platform for translational cancer research. SIGNIFICANCE: Patient-derived xenografts of lung cancer retain key features observed in the originating patient tumors and show expected responses to treatment with standard-of-care agents, providing experimentally tractable and reproducible models for preclinical investigations.

REST regulates distinct transcriptional networks in embryonic and neural stem cells.

The maintenance of pluripotency and specification of cellular lineages during embryonic development are controlled by transcriptional regulatory networks, which coordinate specific sets of genes through both activation and repression. The transcriptional repressor RE1-silencing transcription factor (REST) plays important but distinct regulatory roles in embryonic (ESC) and neural (NSC) stem cells. We investigated how these distinct biological roles are effected at a genomic level. We present integrated, comparative genome- and transcriptome-wide analyses of transcriptional networks governed by REST in mouse ESC and NSC. The REST recruitment profile has dual components: a developmentally independent core that is common to ESC, NSC, and differentiated cells; and a large, ESC-specific set of target genes. In ESC, the REST regulatory network is highly integrated into that of pluripotency factors Oct4-Sox2-Nanog. We propose that an extensive, pluripotency-specific recruitment profile lends REST a key role in the maintenance of the ESC phenotype.

Zebrafish whole-adult-organism chemogenomics for large-scale predictive and discovery chemical biology.

The ability to perform large-scale, expression-based chemogenomics on whole adult organisms, as in invertebrate models (worm and fly), is highly desirable for a vertebrate model but its feasibility and potential has not been demonstrated. We performed expression-based chemogenomics on the whole adult organism of a vertebrate model, the zebrafish, and demonstrated its potential for large-scale predictive and discovery chemical biology. Focusing on two classes of compounds with wide implications to human health, polycyclic (halogenated) aromatic hydrocarbons [P(H)AHs] and estrogenic compounds (ECs), we generated robust prediction models that can discriminate compounds of the same class from those of different classes in two large independent experiments. The robust expression signatures led to the identification of biomarkers for potent aryl hydrocarbon receptor (AHR) and estrogen receptor (ER) agonists, respectively, and were validated in multiple targeted tissues. Knowledge-based data mining of human homologs of zebrafish genes revealed highly conserved chemical-induced biological responses/effects, health risks, and novel biological insights associated with AHR and ER that could be inferred to humans. Thus, our study presents an effective, high-throughput strategy of capturing molecular snapshots of chemical-induced biological states of a whole adult vertebrate that provides information on biomarkers of effects, deregulated signaling pathways, and possible affected biological functions, perturbed physiological systems, and increased health risks. These findings place zebrafish in a strategic position to bridge the wide gap between cell-based and rodent models in chemogenomics research and applications, especially in preclinical drug discovery and toxicology.

Transfer learning-trained convolutional neural networks identify novel MRI biomarkers of Alzheimer's disease progression.

INTRODUCTION: Genome-wide association studies (GWAS) for late onset Alzheimer's disease (AD) may miss genetic variants relevant for delineating disease stages when using clinically defined case/control as a phenotype due to its loose definition and heterogeneity. METHODS: We use a transfer learning technique to train three-dimensional convolutional neural network (CNN) models based on structural magnetic resonance imaging (MRI) from the screening stage in the Alzheimer's Disease Neuroimaging Initiative consortium to derive image features that reflect AD progression. RESULTS: CNN-derived image phenotypes are significantly associated with fasting metabolites related to early lipid metabolic changes as well as insulin resistance and with genetic variants mapped to candidate genes enriched for amyloid beta degradation, tau phosphorylation, calcium ion binding-dependent synaptic loss, APP-regulated inflammation response, and insulin resistance. DISCUSSION: This is the first attempt to show that non-invasive MRI biomarkers are linked to AD progression characteristics, reinforcing their use in early AD diagnosis and monitoring.

Genome-wide estimation of firing efficiencies of origins of DNA replication from time-course copy number variation data.

BACKGROUND: DNA replication is a fundamental biological process during S phase of cell division. It is initiated from several hundreds of origins along whole chromosome with different firing efficiencies (or frequency of usage). Direct measurement of origin firing efficiency by techniques such as DNA combing are time-consuming and lack the ability to measure all origins. Recent genome-wide study of DNA replication approximated origin firing efficiency by indirectly measuring other quantities related to replication. However, these approximation methods do not reflect properties of origin firing and may lead to inappropriate estimations. RESULTS: In this paper, we develop a probabilistic model - Spanned Firing Time Model (SFTM) to characterize DNA replication process. The proposed model reflects current understandings about DNA replication. Origins in an individual cell may initiate replication randomly within a time window, but the population average exhibits a temporal program with some origins replicated early and the others late. By estimating DNA origin firing time and fork moving velocity from genome-wide time-course S-phase copy number variation data, we could estimate firing efficiency of all origins. The estimated firing efficiency is correlated well with the previous studies in fission and budding yeasts. CONCLUSIONS: The new probabilistic model enables sensitive identification of origins as well as genome-wide estimation of origin firing efficiency. We have successfully estimated firing efficiencies of all origins in S. cerevisiae, S. pombe and human chromosomes 21 and 22.

The role of post-transcriptional RNA processing and plasmid vector sequences on transient transgene expression in zebrafish.

A tissue-specific transgenic model was employed to test the effects of intron and vector sequences on transgene expression in zebrafish after microinjection. In this model, the 2.3 kb promoter taken from the 5' upstream region of the transcription initiation site of keratin 4 (krt4) was used to drive the enhanced green fluorescence protein (EGFP) reporter gene in a transgenic vector. For assaying the strength of EGFP expression, the effects of including an intron before the EGFP coding region or using different forms of DNA, including circular plasmid, linear full-length plasmid, and the linear transgene coding region without any prokaryotic vector sequence, were tested. After microinjection, the transgene expression was analyzed using transient assays. Consequently, further comparative analysis supported by Fisher's exact test was performed based on the data generated by analyzing the strength of the transgene expression. It was shown that inclusion of an intron in the construct increases the transgene expression in a transient transgenic zebrafish assay. Furthermore, the circular plasmid containing the transgene produced the strongest EGFP expression.

Bioinformatics Core Survey Highlights the Challenges Facing Data Analysis Facilities.

Over the last decade, the cost of -omics data creation has decreased 10-fold, whereas the need for analytical support for those data has increased exponentially. Consequently, bioinformaticians face a second wave of challenges: novel applications of existing approaches (e.g., single-cell RNA sequencing), integration of -omics data sets of differing size and scale (e.g., spatial transcriptomics), as well as novel computational and statistical methods, all of which require more sophisticated pipelines and data management. Nonetheless, bioinformatics cores are often asked to operate under primarily a cost-recovery model, with limited institutional support. Seeing the need to assess bioinformatics core operations, the Association of Biomolecular Resource Facilities Genomics Bioinformatics Research Group conducted a survey to answer questions about staffing, services, financial models, and challenges to better understand the challenges bioinformatics core facilities are currently faced with and will need to address going forward. Of the respondent groups, we chose to focus on the survey data from smaller cores, which made up the majority. Although all cores indicated similar challenges in terms of changing technologies and analysis needs, small cores tended to have the added challenge of funding their operations largely through cost-recovery models with heavy administrative burdens.

CUP-AI-Dx: A tool for inferring cancer tissue of origin and molecular subtype using RNA gene-expression data and artificial intelligence.

BACKGROUND: Cancer of unknown primary (CUP), representing approximately 3-5% of all malignancies, is defined as metastatic cancer where a primary site of origin cannot be found despite a standard diagnostic workup. Because knowledge of a patient's primary cancer remains fundamental to their treatment, CUP patients are significantly disadvantaged and most have a poor survival outcome. Developing robust and accessible diagnostic methods for resolving cancer tissue of origin, therefore, has significant value for CUP patients. METHODS: We developed an RNA-based classifier called CUP-AI-Dx that utilizes a 1D Inception convolutional neural network (1D-Inception) model to infer a tumor's primary tissue of origin. CUP-AI-Dx was trained using the transcriptional profiles of 18,217 primary tumours representing 32 cancer types from The Cancer Genome Atlas project (TCGA) and International Cancer Genome Consortium (ICGC). Gene expression data was ordered by gene chromosomal coordinates as input to the 1D-CNN model, and the model utilizes multiple convolutional kernels with different configurations simultaneously to improve generality. The model was optimized through extensive hyperparameter tuning, including different max-pooling layers and dropout settings. For 11 tumour types, we also developed a random forest model that can classify the tumour's molecular subtype according to prior TCGA studies. The optimised CUP-AI-Dx tissue of origin classifier was tested on 394 metastatic samples from 11 tumour types from TCGA and 92 formalin-fixed paraffin-embedded (FFPE) samples representing 18 cancer types from two clinical laboratories. The CUP-AI-Dx molecular subtype was also independently tested on independent ovarian and breast cancer microarray datasets FINDINGS: CUP-AI-Dx identifies the primary site with an overall top-1-accuracy of 98.54% in cross-validation and 96.70% on a test dataset. When applied to two independent clinical-grade RNA-seq datasets generated from two different institutes from the US and Australia, our model predicted the primary site with a top-1-accuracy of 86.96% and 72.46% respectively. INTERPRETATION: The CUP-AI-Dx predicts tumour primary site and molecular subtype with high accuracy and therefore can be used to assist the diagnostic work-up of cancers of unknown primary or uncertain origin using a common and accessible genomics platform. FUNDING: NIH R35 GM133562, NCI P30 CA034196, Victorian Cancer Agency Australia.

Ablation of phosphoinositide-3-kinase class II alpha suppresses hepatoma cell proliferation.

Cancer such as hepatocellular carcinoma (HCC) is characterized by complex perturbations in multiple signaling pathways, including the phosphoinositide-3-kinase (PI3K/AKT) pathways. Herein we investigated the role of PI3K catalytic isoforms, particularly class II isoforms in HCC proliferation. Among the siRNAs tested against the eight known catalytic PI3K isoforms, specific ablation of class II PI3K alpha (PIK3C2alpha) was the most effective in impairing cell growth and this was accompanied by concomitant decrease in PIK3C2alpha mRNA and protein levels. Colony formation ability of cells deficient for PIK3C2alpha was markedly reduced and growth arrest was associated with increased caspase 3 levels. A small but significant difference in gene dosage and expression levels was detected between tumor and non-tumor tissues in a cohort of 19 HCC patients. Taken together, these data suggest for the first time that in addition to class I PI3Ks in cancer, class II PIK3C2alpha can modulate HCC cell growth.

Genomic data analysis workflows for tumors from patient-derived xenografts (PDXs): challenges and guidelines.

BACKGROUND: Patient-derived xenograft (PDX) models are in vivo models of human cancer that have been used for translational cancer research and therapy selection for individual patients. The Jackson Laboratory (JAX) PDX resource comprises 455 models originating from 34 different primary sites (as of 05/08/2019). The models undergo rigorous quality control and are genomically characterized to identify somatic mutations, copy number alterations, and transcriptional profiles. Bioinformatics workflows for analyzing genomic data obtained from human tumors engrafted in a mouse host (i.e., Patient-Derived Xenografts; PDXs) must address challenges such as discriminating between mouse and human sequence reads and accurately identifying somatic mutations and copy number alterations when paired non-tumor DNA from the patient is not available for comparison. RESULTS: We report here data analysis workflows and guidelines that address these challenges and achieve reliable identification of somatic mutations, copy number alterations, and transcriptomic profiles of tumors from PDX models that lack genomic data from paired non-tumor tissue for comparison. Our workflows incorporate commonly used software and public databases but are tailored to address the specific challenges of PDX genomics data analysis through parameter tuning and customized data filters and result in improved accuracy for the detection of somatic alterations in PDX models. We also report a gene expression-based classifier that can identify EBV-transformed tumors. We validated our analytical approaches using data simulations and demonstrated the overall concordance of the genomic properties of xenograft tumors with data from primary human tumors in The Cancer Genome Atlas (TCGA). CONCLUSIONS: The analysis workflows that we have developed to accurately predict somatic profiles of tumors from PDX models that lack normal tissue for comparison enable the identification of the key oncogenic genomic and expression signatures to support model selection and/or biomarker development in therapeutic studies. A reference implementation of our analysis recommendations is available at https://github.com/TheJacksonLaboratory/PDX-Analysis-Workflows .

Identification of cell cycle-regulated genes in fission yeast.

Cell cycle progression is both regulated and accompanied by periodic changes in the expression levels of a large number of genes. To investigate cell cycle-regulated transcriptional programs in the fission yeast Schizosaccharomyces pombe, we developed a whole-genome oligonucleotide-based DNA microarray. Microarray analysis of both wild-type and cdc25 mutant cell cultures was performed to identify transcripts whose levels oscillated during the cell cycle. Using an unsupervised algorithm, we identified 747 genes that met the criteria for cell cycle-regulated expression. Peaks of gene expression were found to be distributed throughout the entire cell cycle. Furthermore, we found that four promoter motifs exhibited strong association with cell cycle phase-specific expression. Examination of the regulation of MCB motif-containing genes through the perturbation of DNA synthesis control/MCB-binding factor (DSC/MBF)-mediated transcription in arrested synchronous cdc10 mutant cell cultures revealed a subset of functional targets of the DSC/MBF transcription factor complex, as well as certain gene promoter requirements. Finally, we compared our data with those for the budding yeast Saccharomyces cerevisiae and found approximately 140 genes that are cell cycle regulated in both yeasts, suggesting that these genes may play an evolutionarily conserved role in regulation of cell cycle-specific processes. Our complete data sets are available at http://giscompute.gis.a-star.edu.sg/~gisljh/CDC.

High-resolution deconstruction of evolution induced by chemotherapy treatments in breast cancer xenografts.

The processes by which tumors evolve are essential to the efficacy of treatment, but quantitative understanding of intratumoral dynamics has been limited. Although intratumoral heterogeneity is common, quantification of evolution is difficult from clinical samples because treatment replicates cannot be performed and because matched serial samples are infrequently available. To circumvent these problems we derived and assayed large sets of human triple-negative breast cancer xenografts and cell cultures from two patients, including 86 xenografts from cyclophosphamide, doxorubicin, cisplatin, docetaxel, or vehicle treatment cohorts as well as 45 related cell cultures. We assayed these samples via exome-seq and/or high-resolution droplet digital PCR, allowing us to distinguish complex therapy-induced selection and drift processes among endogenous cancer subclones with cellularity uncertainty <3%. For one patient, we discovered two predominant subclones that were granularly intermixed in all 48 co-derived xenograft samples. These two subclones exhibited differential chemotherapy sensitivity-when xenografts were treated with cisplatin for 3 weeks, the post-treatment volume change was proportional to the post-treatment ratio of subclones on a xenograft-to-xenograft basis. A subsequent cohort in which xenografts were treated with cisplatin, allowed a drug holiday, then treated a second time continued to exhibit this proportionality. In contrast, xenografts from other treatment cohorts, spatially dissected xenograft fragments, and cell cultures evolved in diverse ways but with substantial population bottlenecks. These results show that ecosystems susceptible to successive retreatment can arise spontaneously in breast cancer in spite of a background of irregular subclonal bottlenecks, and our work provides to our knowledge the first quantification of the population genetics of such a system. Intriguingly, in such an ecosystem the ratio of common subclones is predictive of the state of treatment susceptibility, showing how measurements of subclonal heterogeneity could guide treatment for some patients.

Adaptive expression responses in the Pol-gamma null strain of S. pombe depleted of mitochondrial genome.

BACKGROUND: DNA polymerase gamma(Pol-gamma) has been shown to be essential for maintenance of the mitochondrial genome (mtDNA) in the petite-positive budding yeast Saccharomyces cerevisiae. Budding yeast cells lacking mitochondria exhibit a slow-growing or petite-colony phenotype. Petite strains fail to grow on non-fermentable carbon sources. However, it is not clear whether the Pol-gamma is required for mtDNA maintenance in the petite-negative fission yeast Schizosaccharomyces pombe. RESULTS: We show that disruption of the nuclear gene pog1+ that encodes Pol-gamma is sufficient to deplete mtDNA in S. pombe. Cells bearing pog1Delta allele require substantial growth periods to form petite colonies. Mitotracker assays indicate that pog1Delta cells are defective in mitochondrial function and EM analyses suggest that pog1Delta cells lack normal mitochondrial structures. Depletion of mtDNA in pog1Delta cells is evident from quantitative real-time PCR assays. Genome-wide expression profiles of pog1Delta and other mtDNA-less cells reveal that many genes involved in response to stimulus, energy derivation by oxidation of organic compounds, cellular carbohydrate metabolism, and energy reserve metabolism are induced. Conversely, many genes encoding proteins involved in amino acid metabolism and oxidative phosphorylation are repressed. CONCLUSION: By showing that Pol-gamma is essential for mtDNA maintenance and disruption of pog1+ alters the genome-wide expression profiles, we demonstrated that cells lacking mtDNA exhibit adaptive nuclear gene expression responses in the petite-negative S. pombe.

multiClust: An R-package for Identifying Biologically Relevant Clusters in Cancer Transcriptome Profiles.

Clustering is carried out to identify patterns in transcriptomics profiles to determine clinically relevant subgroups of patients. Feature (gene) selection is a critical and an integral part of the process. Currently, there are many feature selection and clustering methods to identify the relevant genes and perform clustering of samples. However, choosing an appropriate methodology is difficult. In addition, extensive feature selection methods have not been supported by the available packages. Hence, we developed an integrative R-package called multiClust that allows researchers to experiment with the choice of combination of methods for gene selection and clustering with ease. Using multiClust, we identified the best performing clustering methodology in the context of clinical outcome. Our observations demonstrate that simple methods such as variance-based ranking perform well on the majority of data sets, provided that the appropriate number of genes is selected. However, different gene ranking and selection methods remain relevant as no methodology works for all studies.

Identifying driver genes in cancer by triangulating gene expression, gene location, and survival data.

Driver genes are directly responsible for oncogenesis and identifying them is essential in order to fully understand the mechanisms of cancer. However, it is difficult to delineate them from the larger pool of genes that are deregulated in cancer (ie, passenger genes). In order to address this problem, we developed an approach called TRIAngulating Gene Expression (TRIAGE through clinico-genomic intersects). Here, we present a refinement of this approach incorporating a new scoring methodology to identify putative driver genes that are deregulated in cancer. TRIAGE triangulates - or integrates - three levels of information: gene expression, gene location, and patient survival. First, TRIAGE identifies regions of deregulated expression (ie, expression footprints) by deriving a newly established measure called the Local Singular Value Decomposition (LSVD) score for each locus. Driver genes are then distinguished from passenger genes using dual survival analyses. Incorporating measurements of gene expression and weighting them according to the LSVD weight of each tumor, these analyses are performed using the genes located in significant expression footprints. Here, we first use simulated data to characterize the newly established LSVD score. We then present the results of our application of this refined version of TRIAGE to gene expression data from five cancer types. This refined version of TRIAGE not only allowed us to identify known prominent driver genes, such as MMP1, IL8, and COL1A2, but it also led us to identify several novel ones. These results illustrate that TRIAGE complements existing tools, allows for the identification of genes that drive cancer and could perhaps elucidate potential future targets of novel anticancer therapeutics.