Scientific knowledge is possible with small-sample classification

A typical small-sample biomarker classification paper discriminates between types of pathology based on, say, 30,000 genes and a small labeled sample of less than 100 points. Some classification rule is used to design the classifier from this data, but we are given no good reason or conditions under which this algorithm should perform well. An error estimation rule is used to estimate the classification error on the population using the same data, but once again we are given no good reason or conditions under which this error estimator should produce a good estimate, and thus we do not know how well the classifier should be expected to perform. In fact, virtually, in all such papers the error estimate is expected to be highly inaccurate. In short, we are given no justification for any claims.Given the ubiquity of vacuous small-sample classification papers in the literature, one could easily conclude that scientific knowledge is impossible in small-sample settings. It is not that thousands of papers overtly claim that scientific knowledge is impossible in regard to their content; rather, it is that they utilize methods that preclude scientific knowledge. In this paper, we argue to the contrary that scientific knowledge in small-sample classification is possible provided there is sufficient prior knowledge. A natural way to proceed, discussed herein, is via a paradigm for pattern recognition in which we incorporate prior knowledge in the whole classification procedure (classifier design and error estimation), optimize each step of the procedure given available information, and obtain theoretical measures of performance for both classifiers and error estimators, the latter being the critical epistemological issue. In sum, we can achieve scientific validation for a proposed small-sample classifier and its error estimate.     

Phosphoproteome analysis of the MAPK pathway reveals previously undetected feedback mechanisms

The RAS‐RAF‐MEK‐ERK (MAPK) pathway is prevalently perturbed in cancer. Recent large‐scale sequencing initiatives profiled thousands of tumors providing insight into alterations at the DNA and RNA levels. These efforts confirmed that key nodes of the MAPK pathway, in particular KRAS and BRAF, are among the most frequently altered proteins in cancer. The establishment of targeted therapies, however, has proven difficult. To decipher the underlying challenges, it is essential to decrypt the phosphorylation network spanned by the MAPK core axis. Using mass spectrometry we identified 2241 phosphorylation sites on 1020 proteins, and measured their responses to inhibition of MEK or ERK. Multiple phosphorylation patterns revealed previously undetected feedback, as upstream signaling nodes, including receptor kinases, showed changes at the phosphorylation level. We provide a dataset rich in potential therapeutic targets downstream of the MAPK cascade. By integrating TCGA (The Cancer Genome Atlas) data, we highlight some downstream phosphoproteins that are frequently altered in cancer. All MS data have been deposited in the ProteomeXchange with identifier PXD003908 ( http://proteomecentral.proteomexchange.org/dataset/PXD003908 ).     

Use of a Florida Gulf Coast Barrier Island by Spring Trans-Gulf Migrants and the Projected Effects of Sea Level Rise on Habitat Availability

Barrier islands on the north coast of the Gulf of Mexico are an internationally important coastal resource. Each spring hundreds of thousands of Nearctic-Neotropical songbirds crossing the Gulf of Mexico during spring migration use these islands because they provide the first landfall for individuals following a trans-Gulf migratory route. The effects of climate change, particularly sea level rise, may negatively impact habitat availability for migrants on barrier islands. Our objectives were (1) to confirm the use of St. George Island, Florida by trans-Gulf migrants and (2) to determine whether forested stopover habitat will be available for migrants on St. George Island following sea level rise. We used avian transect data, geographic information systems, remote sensing, and simulation modelling to investigate the potential effects of three different sea level rise scenarios (0.28 m, 0.82 m, and 2 m) on habitat availability for trans-Gulf migrants. We found considerable use of the island by spring trans-Gulf migrants. Migrants were most abundant in areas with low elevation, high canopy height, and high coverage of forests and scrub/shrub. A substantial percentage of forest (44%) will be lost by 2100 assuming moderate sea level rise (0.82 m). Thus, as sea level rise progresses, less forests will be available for migrants during stopover. Many migratory bird species’ populations are declining, and degradation of barrier island stopover habitat may further increase the cost of migration for many individuals. To preserve this coastal resource, conservation and wise management of migratory stopover areas, especially near ecological barriers like the Gulf of Mexico, will be essential as sea levels rise.     

Drugging the undruggable: gabapentin,pregabalin and the calcium channel α2δ subunit

In the post-genomic era, the idea of using the sequence of a protein to determine its potential role as a drug target has gained currency. The goal of this approach to drug discovery is to use the sequence of a protein that is known to bind a specific ligand or drug, along with the known structure of the ligand binding site, to predict other similar proteins that are also “druggable”. Gabapentin (Neurontin) and pregabalin (Lyrica) are drugs currently in the clinic that were developed based on the hypothesis that generating non-hydrolyzable analogs of GABA would lead to the development of antiepileptic agents. While these compounds are indeed good anticonvulsants, their activity is surprisingly not due to activity in the GABAergic system. By purifying the protein to which gabapentin bound, and determining its identity as the α₂δ₁ subunit of voltage gated calcium channels, it was possible to make progress in developing new compounds with similar activities to gabapentin, including pregabalin. The recognition of the α₂δ₁ subunit as the receptor for these drugs also meant that related proteins, such as α₂δ₃, may be interesting targets for novel pain therapeutics.     

Microbial enrichment and multiplexed microfiltration for accelerated detection of Salmonella in spinach

To attain Salmonella detection thresholds in spinach suspensions using enrichment media requires at least 24 hr. Separation and concentration of selected microorganisms via microfiltration and microfugation reduce time for sample preparation, especially when working with large volumes of vegetable suspensions. This facilitates accelerated detection of Salmonella in spinach suspensions, and may contribute to effectively monitoring this pathogen before it reaches the consumer. We report a microfiltration-based protocol for accelerated sample preparation to concentrate and recover ≤1 colony forming unit (CFU) Salmonella/g pathogen-free spinach. Store-bought samples of spinach and a spinach plant subjected to two environmental conditions (temperature and light exposure) during its production were tested. The overall procedure involves extraction with buffer, a short enrichment step, prefiltration using a nylon filter, crossflow hollow fiber microfiltration, and retentate centrifugation to bring microbial cells to detection levels. Based on 1 CFU Salmonella/g frozen spinach, and a Poisson distribution statistical analyses with 99% probability, we calculated that 3 hr of incubation, when followed by microfiltration, is sufficient to reach the 2 log concentration required for Salmonella detection within 7 hr. Longer enrichment times (5 hr or more) is needed for concentrations lower than 1 CFU Salmonella/g of ready to eat spinach. The recovered microbial cells were identified and confirmed as Salmonella using both polymerase chain reaction (PCR) and plating methods. Different environmental conditions tested during production did not affect Salmonella viability; this demonstrated the broad adaptability of Salmonella and emphasized the need for methods that enable efficient monitoring of production for the presence of this pathogen.     

Detoxification of olefinic epoxides and nucleotide excision repair of epoxide-mediated DNA damage: Insights from animal models examining human sensitivity to 1,3-butadiene

1,3-Butadiene (BD) is a well-documented mutagen and carcinogen in rodents and is currently classified as a probable carcinogen in humans. Studies investigating workers exposed to BD indicate that, in some plants, there may be an increased genetic risk, and that polymorphisms in biotransformation and DNA repair proteins may modulate genetic susceptibility. To investigate the role of genetic polymorphisms in microsomal epoxide hydrolase (mEH) or nucleotide excision repair (NER) in contributing to the mutagenicity of BD, we conducted a series of experiments in which mice lacking mEH or NER activity were exposed to BD by inhalation or to the reactive epoxide metabolites of BD (epoxybutene-EB or diepoxybutane-DEB) by i.p. injection. Genetic susceptibility was measured using the Hprt cloning assay. Both deficient strains of mouse were significantly more sensitive to the mutagenic effects of BD and the injected epoxides. These studies provide support for the critical role that mEH plays in the biotransformation of BD, and the role that NER plays in maintaining genomic integrity following exposure to BD. Additional studies are needed to examine the importance of base excision repair (BER) in maintaining genomic integrity, the differential formation of DNA and protein adducts in deficient strains, and the potential for enhanced sensitivity to BD genotoxicity in mice either lacking or deficient in both biotransformation and DNA repair activity.     

Recurrent locus-specific mutation resulting from a cryptic ectopic insertion in Neurospora

New mutations are found among approximately 20% of progeny when one or both parents carry eas allele UCLA191 (eas(UCLA), easily wettable, hydrophobin-deficient, linkage group II). The mutations inactivate the wild-type allele of cya-8 (cytochrome aa3 deficient, linkage group VII), resulting in thin, "transparent" mycelial growth. Other eas alleles fail to produce cya-8 mutant progeny. The recurrent cya-8 mutations are attributed to repeat-induced point mutation (RIP) resulting from a duplicated copy of cya-8+ that was inserted ectopically at eas when the UCLA191 mutation occurred. As expected for RIP, eas(UCLA)-induced cya-8 mutations occur during nuclear proliferation prior to karyogamy. When only one parent is eas(UCLA), the new mutations arise exclusively in eas(UCLA) nuclei. Mutation of cya-8 is suppressed when a long unlinked duplication is present. Stable cya-8 mutations are effectively eliminated in crosses homozygous for rid, a recessive suppressor of RIP. The eas(UCLA) allele is associated with a long paracentric inversion. A discontinuity is present in eas(UCLA) DNA. The eas promoter is methylated in cya-8 progeny of eas(UCLA), presumably by the spreading of methylation beyond the adjoining RIP-inactivated duplication. These findings support a model in which an ectopic insertion that created a mutation at the target site acts as a locus-specific mutator via RIP.