The Marker State Space (MSS) Method for Classifying Clinical Samples

The development of accurate clinical biomarkers has been challenging in part due to the diversity between patients and diseases. One approach to account for the diversity is to use multiple markers to classify patients, based on the concept that each individual marker contributes information from its respective subclass of patients. Here we present a new strategy for developing biomarker panels that accounts for completely distinct patient subclasses. Marker State Space (MSS) defines “marker states” based on all possible patterns of high and low values among a panel of markers. Each marker state is defined as either a case state or a control state, and a sample is classified as case or control based on the state it occupies. MSS was used to define multi-marker panels that were robust in cross validation and training-set/test-set analyses and that yielded similar classification accuracy to several other classification algorithms. A three-marker panel for discriminating pancreatic cancer patients from control subjects revealed subclasses of patients based on distinct marker states. MSS provides a straightforward approach for modeling highly divergent subclasses of patients, which may be adaptable for diverse applications.     

A comparison of the structure and properties of normal human transferrin and a genetic variant of human transferrin

1. A rare genetic variant of human serum transferrin (TfBSHAW) is reported. 2. The variant and normal transferrins have been purified. 3. The two proteins have been shown to be identical with respect to their molecular weights, heat stability, iron uptake and absorbance spectra. 4. The amino acid substitution is thought to be isoleucine replaced by asparagine at either position 378 or position 381. 5. The ferric iron bound to the C-site of TfBSHAW is unstable in the presence of protons or 6 M urea.     

Screening for positive allosteric modulators: assessment of modulator concentration-response curves as a screening paradigm

Targeting allosteric binding sites represents a powerful mechanism for selectively modulating receptor function. The advent of functional assays as the screening method of choice is leading to an increase in the number of allosteric modulators identified. These include positive allosteric modulators that can increase the affinity of the orthosteric agonist and potentiate the evoked response. A common method for screening for positive allosteric modulators is to examine a concentration-response (C/R) curve to the putative modulator in the presence of a single, low concentration of agonist. The study reported here has used data simulations for positive allosteric modulators according to the allosteric ternary complex model to generate modulator C/R curves. The results are then compared to the mechanistic parameters used to simulate the data. It is clear from the simulations that the potency of a positive modulator C/R curve in a screening assay is the product of both its affinity and positive cooperativity. However, it is often difficult to tell which parameter dominates the response; not knowing the actual affinity or cooperativity of a ligand may have consequences for receptor selectivity. Further modeling demonstrates that the use and choice of single agonist concentration, as well as changes in the agonist curve Hill slope, can have significant effects on the modulator C/R curve. Finally, the quantitative relationship between modulator C/R curves and the allosteric ternary complex model is explored. These simulations emphasize the importance of careful interpretation of screening data and of conducting full mechanism of action studies for positive allosteric modulators.     

Learning generative models for protein fold families

We introduce a new approach to learning statistical models from multiple sequence alignments (MSA) of proteins. Our method, called GREMLIN (Generative REgularized ModeLs of proteINs), learns an undirected probabilistic graphical model of the amino acid composition within the MSA. The resulting model encodes both the position-specific conservation statistics and the correlated mutation statistics between sequential and long-range pairs of residues. Existing techniques for learning graphical models from MSA either make strong, and often inappropriate assumptions about the conditional independencies within the MSA (e.g., Hidden Markov Models), or else use suboptimal algorithms to learn the parameters of the model. In contrast, GREMLIN makes no a priori assumptions about the conditional independencies within the MSA. We formulate and solve a convex optimization problem, thus guaranteeing that we find a globally optimal model at convergence. The resulting model is also generative, allowing for the design of new protein sequences that have the same statistical properties as those in the MSA. We perform a detailed analysis of covariation statistics on the extensively studied WW and PDZ domains and show that our method out-performs an existing algorithm for learning undirected probabilistic graphical models from MSA. We then apply our approach to 71 additional families from the PFAM database and demonstrate that the resulting models significantly out-perform Hidden Markov Models in terms of predictive accuracy.     

Neurochemical Changes in LPA<Subscript>1</Subscript> Receptor Deficient Mice – A Putative Model of Schizophrenia

LPA₁ is a G_i-coupled seven transmembrane receptor with high affinity for the ligand lysophosphatidic acid. We have investigated the effect of targeted deletion at the lpa₁ locus on evoked release of amino acids from hippocampal slices, using in vitro superfusion techniques, and evoked 5-HT efflux from the dorsal raphe nucleus, using in vitro fast cyclic voltammetry. Superfusion of hippocampal slices revealed that basal levels of tyrosine, aspartate and glutamate release were significantly increased while K⁺-evoked release of glutamate and GABA were significantly decreased in lpa₁^(–/–) mice. Fast cyclic voltammetry measurements in the dorsal raphe nucleus demonstrated significant decreases in electrically evoked 5-HT efflux in lpa₁^(–/–) mice. In summary, these data demonstrate that the lpa₁ mutation produces a number of changes in neurotransmitters that have been associated with a schizophrenic-like pathology.     

Extensive gaps and biases in our knowledge of a well‐known fauna: implications for integrating biological traits into macroecology

Aim Ecologists seeking to describe patterns at ever larger scales require compilations of data on the global abundance and distribution of species. Comparable compilations of biological data are needed to elucidate the mechanisms behind these patterns, but have received far less attention. We assess the availability of biological data across an entire assemblage: the well‐documented demersal marine fauna of the United Kingdom. We also test whether data availability for a species depends on its taxonomic group, maximum body size, the number of times it has been recorded in a global biogeographic database, or its commercial and conservation importance. Location Seas of the United Kingdom. Methods We defined a demersal marine fauna of 973 species from 15 phyla and 40 classes using five extensive surveys around the British Isles. We then quantified the availability of data on eight key biological traits (termed biological knowledge) for each species from online databases. Relationships between biological knowledge and our predictors were tested with generalized linear models. Results Full data on eight fundamental biological traits exist for only 9% (n= 88) of the UK demersal marine fauna, and 20% of species completely lack data. Clear trends in our knowledge exist: fish (median biological knowledge score = six traits) are much better known than invertebrates (one trait). Biological knowledge increases with biogeographic knowledge and (to a lesser extent) with body size, and is greater in species that are commercially exploited or of conservation concern. Main conclusions Our analysis reveals deep ignorance of the basic biology of a well‐studied fauna, highlighting the need for far greater efforts to compile biological trait data. Clear biases in our knowledge, relating to how well sampled or ‘important’ species are suggests that caution is required in extrapolating small subsets of biologically well‐known species to ecosystem‐level studies.     

Learning generative models of molecular dynamics

We introduce three algorithms for learning generative models of molecular structures from molecular dynamics simulations. The first algorithm learns a Bayesian-optimal undirected probabilistic model over user-specified covariates (e.g., fluctuations, distances, angles, etc). L1 regularization is used to ensure sparse models and thus reduce the risk of over-fitting the data. The topology of the resulting model reveals important couplings between different parts of the protein, thus aiding in the analysis of molecular motions. The generative nature of the model makes it well-suited to making predictions about the global effects of local structural changes (e.g., the binding of an allosteric regulator). Additionally, the model can be used to sample new conformations. The second algorithm learns a time-varying graphical model where the topology and parameters change smoothly along the trajectory, revealing the conformational sub-states. The last algorithm learns a Markov Chain over undirected graphical models which can be used to study and simulate kinetics. We demonstrate our algorithms on multiple molecular dynamics trajectories.     

Monoclonal anti-β1-adrenergic receptor antibodies activate G protein signaling in the absence of β-arrestin recruitment

Thermostabilized G protein-coupled receptors used as antigens for in vivo immunization have resulted in the generation of functional agonistic anti-β₁-adrenergic (β₁AR) receptor monoclonal antibodies (mAbs). The focus of this study was to examine the pharmacology of these antibodies to evaluate their mechanistic activity at β₁AR. Immunization with the β₁AR stabilized receptor yielded five stable hybridoma clones, four of which expressed functional IgG, as determined in cell-based assays used to evaluate cAMP stimulation. The antibodies bind diverse epitopes associated with low nanomolar agonist activity at β₁AR, and they appeared to show some degree of biased signaling as they were inactive in an assay measuring signaling through β-arrestin. In vitro characterization also verified different antibody-receptor interactions reflecting the different epitopes on the extracellular surface of β₁AR to which the mAbs bind. The anti-β₁AR mAbs only demonstrated agonist activity when in dimeric antibody format, but not as the monomeric Fab format, suggesting that agonist activation may be mediated through promoting receptor dimerization. Finally, we have also shown that at least one of these antibodies exhibits in vivo functional activity at a therapeutically-relevant dose producing an increase in heart rate consistent with β₁AR agonism.     

Searching for SNPs with cloud computing

As DNA sequencing outpaces improvements in computer speed, there is a critical need to accelerate tasks like alignment and SNP calling. Crossbow is a cloud-computing software tool that combines the aligner Bowtie and the SNP caller SOAPsnp. Executing in parallel using Hadoop, Crossbow analyzes data comprising 38-fold coverage of the human genome in three hours using a 320-CPU cluster rented from a cloud computing service for about $85. Crossbow is available from .     

Ultrafast and memory-efficient alignment of short DNA sequences to the human genome

Bowtie is an ultrafast, memory-efficient alignment program for aligning short DNA sequence reads to large genomes. For the human genome, Burrows-Wheeler indexing allows Bowtie to align more than 25 million reads per CPU hour with a memory footprint of approximately 1.3 gigabytes. Bowtie extends previous Burrows-Wheeler techniques with a novel quality-aware backtracking algorithm that permits mismatches. Multiple processor cores can be used simultaneously to achieve even greater alignment speeds. Bowtie is open source .