Completed genome and emergence scenario of the multidrug-resistant nosocomial pathogen Staphylococcus epidermidis ST215.

BACKGROUND: A multidrug-resistant lineage of Staphylococcus epidermidis named ST215 is a common cause of prosthetic joint infections and other deep surgical site infections in Northern Europe, but is not present elsewhere. The increasing resistance among S. epidermidis strains is a global concern. We used whole-genome sequencing to characterize ST215 from healthcare settings. RESULTS: We completed the genome of a ST215 isolate from a Swedish hospital using short and long reads, resulting in a circular 2,676,787 bp chromosome and a 2,326 bp plasmid. The new ST215 genome was placed in phylogenetic context using 1,361 finished public S. epidermidis reference genomes. We generated 10 additional short-read ST215 genomes and 11 short-read genomes of ST2, which is another common multidrug-resistant lineage at the same hospital. We studied recombination's role in the evolution of ST2 and ST215, and found multiple recombination events averaging 30-50 kb. By comparing the results of antimicrobial susceptibility testing for 31 antimicrobial drugs with the genome content encoding antimicrobial resistance in the ST215 and ST2 isolates, we found highly similar resistance traits between the isolates, with 22 resistance genes being shared between all the ST215 and ST2 genomes. The ST215 genome contained 29 genes that were historically identified as virulence genes of S. epidermidis ST2. We established that in the nucleotide sequence stretches identified as recombination events, virulence genes were overrepresented in ST215, while antibiotic resistance genes were overrepresented in ST2. CONCLUSIONS: This study features the extensive antibiotic resistance and virulence gene content in ST215 genomes. ST215 and ST2 lineages have similarly evolved, acquiring resistance and virulence through genomic recombination. The results highlight the threat of new multidrug-resistant S. epidermidis lineages emerging in healthcare settings.

A genome-wide test for paternal indirect genetic effects on lifespan in Drosophila melanogaster.

Exposing sires to various environmental manipulations has demonstrated that paternal effects can be non-trivial also in species where male investment in offspring is almost exclusively limited to sperm. Whether paternal effects also have a genetic component (i.e. paternal indirect genetic effects (PIGEs)) in such species is however largely unknown, primarily because of methodological difficulties separating indirect from direct effects of genes. PIGEs may nevertheless be important since they have the capacity to contribute to evolutionary change. Here we use Drosophila genetics to construct a breeding design that allows testing nearly complete haploid genomes (more than 99%) for PIGEs. Using this technique, we estimate the variance in male lifespan due to PIGEs among four populations and compare this to the total paternal genetic variance (the sum of paternal indirect and direct genetic effects). Our results indicate that a substantial part of the total paternal genetic variance results from PIGEs. A screen of 38 haploid genomes, randomly sampled from a single population, suggests that PIGEs also influence variation in lifespan within populations. Collectively, our results demonstrate that PIGEs may constitute an underappreciated source of phenotypic variation.

In Silico Identification of Potential Thyroid Hormone System Disruptors among Chemicals in Human Serum and Chemicals with a High Exposure Index.

Data on toxic effects are at large missing the prevailing understanding of the risks of industrial chemicals. Thyroid hormone (TH) system disruption includes interferences of the life cycle of the thyroid hormones and may occur in various organs. In the current study, high-throughput screening data available for 14 putative molecular initiating events of adverse outcome pathways, related to disruption of the TH system, were used to develop 19 in silico models for identification of potential thyroid hormone system-disrupting chemicals. The conformal prediction framework with the underlying Random Forest was used as a wrapper for the models allowing for setting the desired confidence level and controlling the error rate of predictions. The trained models were then applied to two different databases: (i) an in-house database comprising xenobiotics identified in human blood and ii) currently used chemicals registered in the Swedish Product Register, which have been predicted to have a high exposure index to consumers. The application of these models showed that among currently used chemicals, fewer were overall predicted as active compared to chemicals identified in human blood. Chemicals of specific concern for TH disruption were identified from both databases based on their predicted activity.

Combining epigenetic and clinicopathological variables improves specificity in prognostic prediction in clear cell renal cell carcinoma.

BACKGROUND: Metastasized clear cell renal cell carcinoma (ccRCC) is associated with a poor prognosis. Almost one-third of patients with non-metastatic tumors at diagnosis will later progress with metastatic disease. These patients need to be identified already at diagnosis, to undertake closer follow up and/or adjuvant treatment. Today, clinicopathological variables are used to risk classify patients, but molecular biomarkers are needed to improve risk classification to identify the high-risk patients which will benefit most from modern adjuvant therapies. Interestingly, DNA methylation profiling has emerged as a promising prognostic biomarker in ccRCC. This study aimed to derive a model for prediction of tumor progression after nephrectomy in non-metastatic ccRCC by combining DNA methylation profiling with clinicopathological variables. METHODS: A novel cluster analysis approach (Directed Cluster Analysis) was used to identify molecular biomarkers from genome-wide methylation array data. These novel DNA methylation biomarkers, together with previously identified CpG-site biomarkers and clinicopathological variables, were used to derive predictive classifiers for tumor progression. RESULTS: The "triple classifier" which included both novel and previously identified DNA methylation biomarkers together with clinicopathological variables predicted tumor progression more accurately than the currently used Mayo scoring system, by increasing the specificity from 50% in Mayo to 64% in our triple classifier at 85% fixed sensitivity. The cumulative incidence of progress (pCIP5yr) was 7.5% in low-risk vs 44.7% in high-risk in M0 patients classified by the triple classifier at diagnosis. CONCLUSIONS: The triple classifier panel that combines clinicopathological variables with genome-wide methylation data has the potential to improve specificity in prognosis prediction for patients with non-metastatic ccRCC.

Mutations in collagen, type XVII, alpha 1 (COL17A1) cause epithelial recurrent erosion dystrophy (ERED).

Corneal dystrophies are a clinically and genetically heterogeneous group of inherited disorders that bilaterally affect corneal transparency. They are defined according to the corneal layer affected and by their genetic cause. In this study, we identified a dominantly inherited epithelial recurrent erosion dystrophy (ERED)-like disease that is common in northern Sweden. Whole-exome sequencing resulted in the identification of a novel mutation, c.2816C>T, p.T939I, in the COL17A1 gene, which encodes collagen type XVII alpha 1. The variant segregated with disease in a genealogically expanded pedigree dating back 200 years. We also investigated a unique COL17A1 synonymous variant, c.3156C>T, identified in a previously reported unrelated dominant ERED-like family linked to a locus on chromosome 10q23-q24 encompassing COL17A1. We show that this variant introduces a cryptic donor site resulting in aberrant pre-mRNA splicing and is highly likely to be pathogenic. Bi-allelic COL17A1 mutations have previously been associated with a recessive skin disorder, junctional epidermolysis bullosa, with recurrent corneal erosions being reported in some cases. Our findings implicate presumed gain-of-function COL17A1 mutations causing dominantly inherited ERED and improve understanding of the underlying pathology.

Epidemiology and ecology of tularemia in Sweden, 1984-2012.

The zoonotic disease tularemia is endemic in large areas of the Northern Hemisphere, but research is lacking on patterns of spatial distribution and connections with ecologic factors. To describe the spatial epidemiology of and identify ecologic risk factors for tularemia incidence in Sweden, we analyzed surveillance data collected over 29 years (1984-2012). A total of 4,830 cases were notified, of which 3,524 met all study inclusion criteria. From the first to the second half of the study period, mean incidence increased 10-fold, from 0.26/100,000 persons during 1984-1998 to 2.47/100,000 persons during 1999-2012 (p<0.001). The incidence of tularemia was higher than expected in the boreal and alpine ecologic regions (p<0.001), and incidence was positively correlated with the presence of lakes and rivers (p<0.001). These results provide a comprehensive epidemiologic description of tularemia in Sweden and illustrate that incidence is higher in locations near lakes and rivers.

Flow cytometry analysis using sysmex UF-1000i classifies uropathogens based on bacterial, leukocyte, and erythrocyte counts in urine specimens among patients with urinary tract infections.

Urinary tract infections (UTIs) are the second most common bacterial infection. Urine culture is the gold standard for diagnosis, but new techniques, such as flow cytometry analysis (FCA), have been introduced. The aim of the present study was to evaluate FCA characteristics regarding bacteriuria, leukocyturia, and erythrocyturia in relation to cultured uropathogens in specimens from patients with a suspected UTI. We also wanted to evaluate whether the FCA characteristics can identify uropathogens prior to culture. From a prospective study, 1,587 consecutive urine specimens underwent FCA prior to culture during January and February 2012. Outpatients and inpatients (79.6% and 19.4%, respectively) were included, of whom women represented 67.5%. In total, 620 specimens yielded growth, of which Escherichia coli represented 65%, Enterococcus spp. 8%, Klebsiella spp. 7%, and Staphylococcus spp. 5%. For the uropathogens, the outcome of FCA was compared against the results for specimens with E. coli and those with a negative culture. E. coli had high bacterial (median, 17,914/µl), leukocyte (median, 348/µl), and erythrocyte (median, 23/µl) counts. With the exception of Klebsiella spp., the majority of the uropathogens had considerable or significantly lower bacterial counts than that of E. coli. High leukocyte counts were found in specimens with Staphylococcus aureus, Proteus mirabilis, Pseudomonas aeruginosa, and group C streptococci. Elevated erythrocyte counts were found for P. vulgaris, P. aeruginosa, and group C streptococci, as well as for Staphylococcus saprophyticus. In essence, FCA adds new information about the bacterial, leukocyte, and erythrocyte counts in urine specimens for different uropathogens. Based on FCA characteristics, uropathogens can be classified and identified prior to culture. E. coli and Klebsiella spp. have similar FCA characteristics.

The shared genome is a pervasive constraint on the evolution of sex-biased gene expression.

Males and females share most of their genomes, and differences between the sexes can therefore not evolve through sequence divergence in protein coding genes. Sexual dimorphism is instead restricted to occur through sex-specific expression and splicing of gene products. Evolution of sexual dimorphism through these mechanisms should, however, also be constrained when the sexes share the genetic architecture for regulation of gene expression. Despite these obstacles, sexual dimorphism is prevalent in the animal kingdom and commonly evolves rapidly. Here, we ask whether the genetic architecture of gene expression is plastic and easily molded by sex-specific selection, or if sexual dimorphism evolves rapidly despite pervasive genetic constraint. To address this question, we explore the relationship between the intersexual genetic correlation for gene expression (rMF), which captures how independently genes are regulated in the sexes, and the evolution of sex-biased gene expression. Using transcriptome data from Drosophila melanogaster, we find that most genes have a high rMF and that genes currently exposed to sexually antagonistic selection have a higher average rMF than other genes. We further show that genes with a high rMF have less pronounced sex-biased gene expression than genes with a low rMF within D. melanogaster and that the strength of the rMF in D. melanogaster predicts the degree to which the sex bias of a gene's expression has changed between D. melanogaster and six other species in the Drosophila genus. In sum, our results show that a shared genome constrains both short- and long-term evolution of sexual dimorphism.

Cycles of light and dark co-ordinate reversible colony differentiation in Listeria monocytogenes.

Recently, several light receptors have been identified in non-phototrophic bacteria, but their physiological roles still remain rather elusive. Here we show that colonies of the saprophytic bacterium Listeria monocytogenes undergo synchronized multicellular behaviour on agar plates, in response to oscillating light/dark conditions, giving rise to alternating ring formation (opaque and translucent rings). On agar plates, bacteria from opaque rings survive increased levels of reactive oxygen species (ROS), as well as repeated cycles of light and dark, better than bacteria from translucent rings. The ring formation is strictly dependent on a blue-light receptor, Lmo0799, acting through the stress-sigma factor, σ(B) . A transposon screening identified 48 mutants unable to form rings at alternating light conditions, with several of them showing a decreased σ(B) activity/level. However, some of the tested mutants displayed a varied σ(B) activity depending on which of the two stress conditions tested (light or H(2) O(2) exposure). Intriguingly, the transcriptional regulator PrfA and the virulence factor ActA were shown to be required for ring formation by a mechanism involving activation of σ(B) . All in all, this suggests a distinct pathway for Lmo0799 that converge into a common signalling pathway for σ(B) activation. Our results show that night and day cycles co-ordinate a reversible differentiation of a L. monocytogenes colony at room temperature, by a process synchronized by a blue-light receptor and σ(B) .

Within-population distribution of trimethoprim resistance in Escherichia coli before and after a community-wide intervention on trimethoprim use.

A 2-year prospective intervention on the prescription of trimethoprim reduced the use by 85% in a health care region with 178,000 inhabitants. Here, we performed before-and-after analyses of the within-population distribution of trimethoprim resistance in Escherichia coli. Phylogenetic and population genetic methods were applied to multilocus sequence typing data of 548 consecutively collected E. coli isolates from clinical urinary specimens. Results were analyzed in relation to antibiotic susceptibility and the presence and genomic location of different trimethoprim resistance gene classes. A total of 163 E. coli sequence types (STs) were identified, of which 68 were previously undescribed. The isolates fell into one of three distinct genetic clusters designated BAPS 1 (E. coli phylogroup B2), BAPS 2 (phylogroup A and B1), and BAPS 3 (phylogroup D), each with a similar frequency before and after the intervention. BAPS 2 and BAPS 3 were positively and BAPS 1 was negatively associated with trimethoprim resistance (odds ratios of 1.97, 3.17, and 0.26, respectively). In before-and-after analyses, trimethoprim resistance frequency increased in BAPS 1 and decreased in BAPS 2. Resistance to antibiotics other than trimethoprim increased in BAPS 2. Analysis of the genomic location of different trimethoprim resistance genes in isolates of ST69, ST58, and ST73 identified multiple independent acquisition events in isolates of the same ST. The results show that despite a stable overall resistance frequency in E. coli before and after the intervention, marked within-population changes occurred. A decrease of resistance in one major genetic cluster was masked by a reciprocal increase in another major cluster.

Pr is a member of a restricted class of σ70-dependent promoters that lack a recognizable -10 element.

The Pr promoter is the first verified member of a class of bacterial σ(70)-promoters that only possess a single match to consensus within its -10 element. In its native context, the activity of this promoter determines the ability of Pseudomonas putida CF600 to degrade phenolic compounds, which provides proof-of-principle for the significance of such promoters. Lack of identity within the -10 element leads to non-detection of Pr-like promoters by current search engines, because of their bias for detection of the -10 motif. Here, we report a mutagenesis analysis of Pr that reveals strict sequence requirements for its activity that includes an essential -15 element and preservation of non-consensus bases within its -35 and -10 elements. We found that highly similar promoters control plasmid- and chromosomally- encoded phenol degradative systems in various Pseudomonads. However, using a purpose-designed promoter-search algorithm and activity analysis of potential candidate promoters, no bona fide Pr-like promoter could be found in the entire genome of P. putida KT2440. Hence, Pr-like σ(70)-promoters, which have the potential to be a widely distributed class of previously unrecognized promoters, are in fact highly restricted and remain in a class of their own.

Elongator complex influences telomeric gene silencing and DNA damage response by its role in wobble uridine tRNA modification.

Elongator complex is required for formation of the side chains at position 5 of modified nucleosides 5-carbamoylmethyluridine (ncm5U34), 5-methoxycarbonylmethyluridine (mcm5U34), and 5-methoxycarbonylmethyl-2-thiouridine (mcm5s²U34) at wobble position in tRNA. These modified nucleosides are important for efficient decoding during translation. In a recent publication, Elongator complex was implicated to participate in telomeric gene silencing and DNA damage response by interacting with proliferating cell nuclear antigen (PCNA). Here we show that elevated levels of tRNA(Lys)(s²UUU), tRNA(Gln)(s²UUG), and tRNA(Glu)(s²UUC), which in a wild-type background contain the mcm5s²U nucleoside at position 34, suppress the defects in telomeric gene silencing and DNA damage response observed in the Elongator mutants. We also found that the reported differences in telomeric gene silencing and DNA damage response of various elp3 alleles correlated with the levels of modified nucleosides at U34. Defects in telomeric gene silencing and DNA damage response are also observed in strains with the tuc2Δ mutation, which abolish the formation of the 2-thio group of the mcm5s²U nucleoside in tRNA(Lys)(mcm5s²UUU), tRNA(Gln)(mcm5s²UUG), and tRNA(Glu)(mcm5s²UUC). These observations show that Elongator complex does not directly participate in telomeric gene silencing and DNA damage response, but rather that modified nucleosides at U34 are important for efficient expression of gene products involved in these processes. Consistent with this notion, we found that expression of Sir4, a silent information regulator required for assembly of silent chromatin at telomeres, was decreased in the elp3Δ mutants.

OnPLS integration of transcriptomic, proteomic and metabolomic data shows multi-level oxidative stress responses in the cambium of transgenic hipI- superoxide dismutase Populus plants.

BACKGROUND: Reactive oxygen species (ROS) are involved in the regulation of diverse physiological processes in plants, including various biotic and abiotic stress responses. Thus, oxidative stress tolerance mechanisms in plants are complex, and diverse responses at multiple levels need to be characterized in order to understand them. Here we present system responses to oxidative stress in Populus by integrating data from analyses of the cambial region of wild-type controls and plants expressing high-isoelectric-point superoxide dismutase (hipI-SOD) transcripts in antisense orientation showing a higher production of superoxide. The cambium, a thin cell layer, generates cells that differentiate to form either phloem or xylem and is hypothesized to be a major reason for phenotypic perturbations in the transgenic plants. Data from multiple platforms including transcriptomics (microarray analysis), proteomics (UPLC/QTOF-MS), and metabolomics (GC-TOF/MS, UPLC/MS, and UHPLC-LTQ/MS) were integrated using the most recent development of orthogonal projections to latent structures called OnPLS. OnPLS is a symmetrical multi-block method that does not depend on the order of analysis when more than two blocks are analysed. Significantly affected genes, proteins and metabolites were then visualized in painted pathway diagrams. RESULTS: The main categories that appear to be significantly influenced in the transgenic plants were pathways related to redox regulation, carbon metabolism and protein degradation, e.g. the glycolysis and pentose phosphate pathways (PPP). The results provide system-level information on ROS metabolism and responses to oxidative stress, and indicate that some initial responses to oxidative stress may share common pathways. CONCLUSION: The proposed data evaluation strategy shows an efficient way of compiling complex, multi-platform datasets to obtain significant biological information.

Outbreaks of tularemia in a boreal forest region depends on mosquito prevalence.

BACKGROUND: We aimed to evaluate the potential association of mosquito prevalence in a boreal forest area with transmission of the bacterial disease tularemia to humans, and model the annual variation of disease using local weather data. METHODS: A prediction model for mosquito abundance was built using weather and mosquito catch data. Then a negative binomial regression model based on the predicted mosquito abundance and local weather data was built to predict annual numbers of humans contracting tularemia in Dalarna County, Sweden. RESULTS: Three hundred seventy humans were diagnosed with tularemia between 1981 and 2007, 94% of them during 7 summer outbreaks. Disease transmission was concentrated along rivers in the area. The predicted mosquito abundance was correlated (0.41, P < .05) with the annual number of human cases. The predicted mosquito peaks consistently preceded the median onset time of human tularemia (temporal correlation, 0.76; P < .05). Our final predictive model included 5 environmental variables and identified 6 of the 7 outbreaks. CONCLUSIONS: This work suggests that a high prevalence of mosquitoes in late summer is a prerequisite for outbreaks of tularemia in a tularemia-endemic boreal forest area of Sweden and that environmental variables can be used as risk indicators.

Transcriptional signatures measured in whole blood correlate with protection against tuberculosis in inbred and outbred mice.

Although BCG has been used for almost 100 years to immunize against Mycobacterium tuberculosis, TB remains a global public health threat. Numerous clinical trials are underway studying novel vaccine candidates and strategies to improve or replace BCG, but vaccine development still lacks a well-defined set of immune correlates to predict vaccine-induced protection against tuberculosis. This study aimed to address this gap by examining transcriptional responses to BCG vaccination in C57BL/6 inbred mice, coupled with protection studies using Diversity Outbred mice. We evaluated relative gene expression in blood obtained from vaccinated mice, because blood is easily accessible, and data can be translated to human studies. We first determined that the average peak time after vaccination is 14 days for gene expression of a small subset of immune-related genes in inbred mice. We then performed global transcriptomic analyses using whole blood samples obtained two weeks after mice were vaccinated with BCG. Using comparative bioinformatic analyses and qRT-PCR validation, we developed a working correlate panel of 18 genes that were highly correlated with administration of BCG but not heat-killed BCG. We then tested this gene panel using BCG-vaccinated Diversity Outbred mice and revealed associations between the expression of a subset of genes and disease outcomes after aerosol challenge with M. tuberculosis. These data therefore demonstrate that blood-based transcriptional immune correlates measured within a few weeks after vaccination can be derived to predict protection against M. tuberculosis, even in outbred populations.

Persistence of cell-mediated immunity three decades after vaccination with the live vaccine strain of Francisella tularensis.

The efficacy of many vaccines against intracellular bacteria depends on the generation of cell-mediated immunity, but studies to determine the duration of immunity are usually confounded by re-exposure. The causative agent of tularemia, Francisella tularensis, is rare in most areas and, therefore, tularemia vaccination is an interesting model for studies of the longevity of vaccine-induced cell-mediated immunity. Here, lymphocyte proliferation and cytokine production in response to F. tularensis were assayed in two groups of 16 individuals, vaccinated 1-3 or 27-34 years previously. As compared to naïve individuals, vaccinees of both groups showed higher proliferative responses and, out of 17 cytokines assayed, higher levels of MIP-1ß, IFN-γ, IL-10, and IL-5 in response to recall stimulation. The responses were very similar in the two groups of vaccinees. A statistical model was developed to predict the immune status of the individuals and by use of two parameters, proliferative responses and levels of IFN-γ, 91.1% of the individuals were correctly classified. Using flow cytometry analysis, we demonstrated that during recall stimulation, expression of IFN-γ by CD4(+) CCR7(+) , CD4(+) CD62L(+) , CD8(+) CCR7(+) , and CD8(+) CD62L(+) cells significantly increased in samples from vaccinated donors. In conclusion, cell-mediated immunity was found to persist three decades after tularemia vaccination without evidence of decline.

DNA-repair gene variants are associated with glioblastoma survival.

Patient outcome from glioma may be influenced by germline variation. Considering the importance of DNA repair in cancer biology as well as in response to treatment, we studied the relationship between 1458 SNPs, which captured the majority of the common genetic variation in 136 DNA repair genes, in 138 glioblastoma samples from Sweden and Denmark. We confirmed our findings in an independent cohort of 121 glioblastoma patients from the UK. Our analysis revealed nine SNPs annotating MSH2, RAD51L1 and RECQL4 that were significantly (p < 0.05) associated with glioblastoma survival.

Buffering of segmental and chromosomal aneuploidies in Drosophila melanogaster.

Chromosomal instability, which involves the deletion and duplication of chromosomes or chromosome parts, is a common feature of cancers, and deficiency screens are commonly used to detect genes involved in various biological pathways. However, despite their importance, the effects of deficiencies, duplications, and chromosome losses on the regulation of whole chromosomes and large chromosome domains are largely unknown. Therefore, to explore these effects, we examined expression patterns of genes in several Drosophila deficiency hemizygotes and a duplication hemizygote using microarrays. The results indicate that genes expressed in deficiency hemizygotes are significantly buffered, and that the buffering effect is general rather than being mainly mediated by feedback regulation of individual genes. In addition, differentially expressed genes in haploid condition appear to be generally more strongly buffered than ubiquitously expressed genes in haploid condition, but, among genes present in triploid condition, ubiquitously expressed genes are generally more strongly buffered than differentially expressed genes. Furthermore, we show that the 4th chromosome is compensated in response to dose differences. Our results suggest general mechanisms have evolved that stimulate or repress gene expression of aneuploid regions as appropriate, and on the 4th chromosome of Drosophila this compensation is mediated by Painting of Fourth (POF).

Classification of microarrays; synergistic effects between normalization, gene selection and machine learning.

BACKGROUND: Machine learning is a powerful approach for describing and predicting classes in microarray data. Although several comparative studies have investigated the relative performance of various machine learning methods, these often do not account for the fact that performance (e.g. error rate) is a result of a series of analysis steps of which the most important are data normalization, gene selection and machine learning. RESULTS: In this study, we used seven previously published cancer-related microarray data sets to compare the effects on classification performance of five normalization methods, three gene selection methods with 21 different numbers of selected genes and eight machine learning methods. Performance in term of error rate was rigorously estimated by repeatedly employing a double cross validation approach. Since performance varies greatly between data sets, we devised an analysis method that first compares methods within individual data sets and then visualizes the comparisons across data sets. We discovered both well performing individual methods and synergies between different methods. CONCLUSION: Support Vector Machines with a radial basis kernel, linear kernel or polynomial kernel of degree 2 all performed consistently well across data sets. We show that there is a synergistic relationship between these methods and gene selection based on the T-test and the selection of a relatively high number of genes. Also, we find that these methods benefit significantly from using normalized data, although it is hard to draw general conclusions about the relative performance of different normalization procedures.

Comparison of Methods for Feature Selection in Clustering of High-Dimensional RNA-Sequencing Data to Identify Cancer Subtypes.

Cancer subtype identification is important to facilitate cancer diagnosis and select effective treatments. Clustering of cancer patients based on high-dimensional RNA-sequencing data can be used to detect novel subtypes, but only a subset of the features (e.g., genes) contains information related to the cancer subtype. Therefore, it is reasonable to assume that the clustering should be based on a set of carefully selected features rather than all features. Several feature selection methods have been proposed, but how and when to use these methods are still poorly understood. Thirteen feature selection methods were evaluated on four human cancer data sets, all with known subtypes (gold standards), which were only used for evaluation. The methods were characterized by considering mean expression and standard deviation (SD) of the selected genes, the overlap with other methods and their clustering performance, obtained comparing the clustering result with the gold standard using the adjusted Rand index (ARI). The results were compared to a supervised approach as a positive control and two negative controls in which either a random selection of genes or all genes were included. For all data sets, the best feature selection approach outperformed the negative control and for two data sets the gain was substantial with ARI increasing from (-0.01, 0.39) to (0.66, 0.72), respectively. No feature selection method completely outperformed the others but using the dip-rest statistic to select 1000 genes was overall a good choice. The commonly used approach, where genes with the highest SDs are selected, did not perform well in our study.