A feature selection method for classification within functional genomics experiments based on the proportional overlapping score

Background

Microarray technology, as well as other functional genomics experiments, allow simultaneous measurements of thousands of genes within each sample. Both the prediction accuracy and interpretability of a classifier could be enhanced by performing the classification based only on selected discriminative genes. We propose a statistical method for selecting genes based on overlapping analysis of expression data across classes. This method results in a novel measure, called proportional overlapping score (POS), of a feature’s relevance to a classification task.

Results

We apply POS, along‐with four widely used gene selection methods, to several benchmark gene expression datasets. The experimental results of classification error rates computed using the Random Forest, k Nearest Neighbor and Support Vector Machine classifiers show that POS achieves a better performance.

Conclusions

A novel gene selection method, POS, is proposed. POS analyzes the expressions overlap across classes taking into account the proportions of overlapping samples. It robustly defines a mask for each gene that allows it to minimize the effect of expression outliers. The constructed masks along‐with a novel gene score are exploited to produce the selected subset of genes.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2105-15-274) contains supplementary material, which is available to authorized users.     

Isolation and sequencing of NOP1. A yeast gene encoding a nucleolar protein homologous to a human autoimmune antigen

We have identified the gene for the yeast nucleolar protein p38 and deduced the primary structure of p38 from its sequence. We propose the name NOP1 (nucleolar protein 1) for this gene. NOP1 encodes a 327 amino acid protein of 34,470 daltons and is flanked by potential promoter and polyadenylation sequences. Blot analyses indicate that the mRNA transcribed from NOP1 is approximately 1.3 kilobases in size and that there is one NOP1 gene per haploid genome. The amino-terminal sequence of p38 is homologous with the 31 known amino-terminal residues of the autoimmune antigen fibrillarin, confirming the previously observed similarity between p38 and this mammalian nucleolar protein. Consistent with this, p38 cross-reacts with serum from a patient with the autoimmune disease scleroderma. A putative nuclear localization signal can be identified in p38. Interestingly, a repetitive amino acid sequence motif begins near the amino terminus of p38. This motif is approximately 80 residues long, is rich in glycine and arginine, and shows striking sequence homology to mammalian nucleolins and certain nucleic acid binding proteins.     

Literature mining, ontologies and information visualization for drug repurposing

The immense growth of MEDLINE coupled with the realization that a vast amount of biomedical knowledge is recorded in free-text format, has led to the appearance of a large number of literature mining techniques aiming to extract biomedical terms and their inter-relations from the scientific literature. Ontologies have been extensively utilized in the biomedical domain either as controlled vocabularies or to provide the framework for mapping relations between concepts in biology and medicine. Literature-based approaches and ontologies have been used in the past for the purpose of hypothesis generation in connection with drug discovery. Here, we review the application of literature mining and ontology modeling and traversal to the area of drug repurposing (DR). In recent years, DR has emerged as a noteworthy alternative to the traditional drug development process, in response to the decreased productivity of the biopharmaceutical industry. Thus, systematic approaches to DR have been developed, involving a variety of in silico, genomic and high-throughput screening technologies. Attempts to integrate literature mining with other types of data arising from the use of these technologies as well as visualization tools assisting in the discovery of novel associations between existing drugs and new indications will also be presented.     

Traveling bands of chemotactic bacteria in the context of population growth

A mathematical model for traveling bands of motile and chemotactic bacteria in the presence of cell growth and death is examined. It is found that asymptotic traveling wave solutions exist in the absence of chemotaxis, due to the balance of growth, death and random motility. Thus random motility confers the ecological advantage of population propagation through migration into nutrient-rich regions. The presence of chemotaxis amplifies this advantage by moving more cells into higher nutrient concentration regions, resulting in larger and faster bands. Therefore there seem to be two types of traveling bands that can be attained by chemotactic bacteria in the presence of growth and death: (1) these growth/death/motility bands; and (2) pure chemotactic ‘Keller-Segel'-type bands. Comparison to experimental observations by Chapman in 1973 indicate that the latter seem to be formed. The relationship between these two types of solution is at present uncertain. The growth/death/motility bands may have relevance on longer time or distance scales characteristic of microbial ecological systems.     

Loss of gremlin delays primordial follicle assembly but does not affect female fertility in mice

The transforming growth factor beta (TGFB) protein family is renowned for its diverse roles in developmental biology including reproduction. Gremlin is a member of the differential screening-selected gene aberrative in neuroblastoma (DAN)/cerberus family of bone morphogenetic protein (BMP) antagonists. Recent studies on gremlin focus on its involvement in embryonic skeletal, lung, and kidney development. To define the role of gremlin (Grem1) in female reproduction, we analyzed postnatal folliculogenesis using global and conditional knockout (cKO) mice for gremlin. Grem1(-/-) mice die within 48 h after birth, and ovaries collected from neonatal Grem1(-/-) mice demonstrated reduced oocyte numbers and delayed primordial follicle development. Transplanting Grem1(-/-) neonatal ovaries showed that folliculogenesis proceeded to large antral follicle stage, but Grem1(-/-) ovaries contained corpora lutea-like structures not found in control-transplanted ovaries. However, Grem1 cKO mice had comparable fertility to control mice. These data suggest that gremlin plays a previously uncharacterized role in the regulation of oocyte numbers and the timing of primordial follicle development, but either it is not required for later folliculogenesis or its loss is possibly compensated by other BMP antagonists.     

Embryoid bodies from mouse stem cells express oxytocin receptor, Oct-4 and DAZL

Oxytocin is a nine amino acid peptide involved in a wide spectrum of physiological functions; predominantly those concerning reproduction and differentiation are of interest. Oxytocin receptors are expressed at early developmental stages of mammals, suggesting that oxytocin might be involved in the determination of the germ stem cell line, at the very early stages of mammalian development. In this respect, the proximate aim of the present study was to confirm and further analyze the existence of oxytocin receptors at a very early level of cell commitment, that is, the determination of germ cells derived from embryoid bodies. To achieve our purpose we have cultured mouse embryonic stem cells under conditions inducing formation of embryoid bodies. In this work, ES cells were allowed to aggregate in a novel medium, “Stefanidis medium” from day 0 to day 14 until formed EBs. RNA was isolated from EBs and using RT-PCR we showed that EBs expressed Oct-4, OTR, OT, and DAZL. To demonstrate simultaneous expression immunocytochemistry was preformed, in which EBs showed strong immunoreactivity for both, OTR and DAZL molecular markers. We found that 35% of the cells displayed OTR, using flow cytometry. In addition, this novel medium showed to increase OTR mRNA. We propose, that at least in murine induced embryoid bodies there is simultaneous expression of oxytocin receptors and germ cell markers (DAZL) in many cells (expressing Oct-4). We thus conclude that, the oxytocin might indeed be a molecule playing a leading role in germ cell determination.     

Phylogenetic analysis of murine leukemia virus sequences from longitudinally sampled chronic fatigue syndrome patients suggests PCR contamination rather than viral evolution

Xenotropic murine leukemia virus (MLV)-related virus (XMRV) has been amplified from human prostate cancer and chronic fatigue syndrome (CFS) patient samples. Other studies failed to replicate these findings and suggested PCR contamination with a prostate cancer cell line, 22Rv1, as a likely source. MLV-like sequences have also been detected in CFS patients in longitudinal samples 15 years apart. Here, we tested whether sequence data from these samples are consistent with viral evolution. Our phylogenetic analyses strongly reject a model of within-patient evolution and demonstrate that the sequences from the first and second time points represent distinct endogenous murine retroviruses, suggesting contamination.