A Comparative Classification Analysis of Abdominal Aortic Aneurysms by Machine Learning Algorithms.

The objective of this work was to perform image-based classification of abdominal aortic aneurysms (AAA) based on their demographic, geometric, and biomechanical attributes. We retrospectively reviewed existing demographics and abdominal computed tomography angiography images of 100 asymptomatic and 50 symptomatic AAA patients who received an elective or emergent repair, respectively, within 1-6 months of their last follow up. An in-house script developed within the MATLAB computational platform was used to segment the clinical images, calculate 53 descriptors of AAA geometry, and generate volume meshes suitable for finite element analysis (FEA). Using a third party FEA solver, four biomechanical markers were calculated from the wall stress distributions. Eight machine learning algorithms (MLA) were used to develop classification models based on the discriminatory potential of the demographic, geometric, and biomechanical variables. The overall classification performance of the algorithms was assessed by the accuracy, area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and precision of their predictions. The generalized additive model (GAM) was found to have the highest accuracy (87%), AUC (89%), and sensitivity (78%), and the third highest specificity (92%), in classifying the individual AAA as either asymptomatic or symptomatic. The k-nearest neighbor classifier yielded the highest specificity (96%). GAM used seven markers (six geometric and one biomechanical) to develop the classifier. The maximum transverse dimension, the average wall thickness at the maximum diameter, and the spatially averaged wall stress were found to be the most influential markers in the classification analysis. A second classification analysis revealed that using maximum diameter alone results in a lower accuracy (79%) than using GAM with seven geometric and biomechanical markers. We infer from these results that biomechanical and geometric measures by themselves are not sufficient to discriminate adequately between population samples of asymptomatic and symptomatic AAA, whereas MLA offer a statistical approach to stratification of rupture risk by combining demographic, geometric, and biomechanical attributes of patient-specific AAA.

Phasic Dopamine Transmission Reflects Initiation Vigor and Exerted Effort in an Action- and Region-Specific Manner.

Initiating a reward-seeking behavior involves deciding on an action, how fast to initiate the action (initiation vigor), as well as how much effort to exert. These processes are thought to involve the mesolimbic dopamine system. Dopamine levels in the ventral striatum rise before initiating a reliably reinforced behavior. However, it is unknown whether dopamine is similarly involved with unreinforced actions (inactive lever presses, premature food port entries, insufficient number of active lever presses). Furthermore, does the dopamine response when initiating an action reflect specific aspects of motivated behavior, such as initiation vigor and exerted effort? Here, we analyzed voltammetry recordings of dopamine levels in the nucleus accumbens (NAcc) core and shell in rats working for food under a progressive ratio reinforcement schedule. We examined dopamine levels when rats initiated distinct actions (active lever presses, inactive lever presses, food port entries) that were temporally separated from cue- and reward-evoked dopamine release. Active lever pressing bouts were preceded by elevated dopamine release in the NAcc shell, as well as in the NAcc core, although only when rats exhibited high initiation vigor. Dopamine levels were transiently reduced in the NAcc core following an unreinforced food port entry and were unchanged throughout the NAcc when initiating inactive lever presses. The effort exerted and vigor to initiate a bout of active lever presses were signaled by dopamine transmission in the NAcc core, but not in the NAcc shell. These results demonstrate that the dopamine response when initiating a behavior is both region- and action-specific. SIGNIFICANCE STATEMENT: Exogenous activation of the mesolimbic dopamine system facilitates motivated behavior. However, a direct relationship has not been established between endogenous phasic dopamine transmission and measures of motivation, such as the vigor to initiate an action and the effort exerted in a bout of activity. The present work demonstrates that the dopamine response when initiating an action depends both upon where dopamine is released and what action is performed. Furthermore, dopamine reflects measures of motivated behavior selectively within the nucleus accumbens core.

Chronic loss of noradrenergic tone produces ß-arrestin2-mediated cocaine hypersensitivity and alters cellular D2 responses in the nucleus accumbens.

Cocaine blocks plasma membrane monoamine transporters and increases extracellular levels of dopamine (DA), norepinephrine (NE) and serotonin (5-HT). The addictive properties of cocaine are mediated primarily by DA, while NE and 5-HT play modulatory roles. Chronic inhibition of dopamine ß-hydroxylase (DBH), which converts DA to NE, increases the aversive effects of cocaine and reduces cocaine use in humans, and produces behavioral hypersensitivity to cocaine and D2 agonism in rodents, but the underlying mechanism is unknown. We found a decrease in ß-arrestin2 (ßArr2) in the nucleus accumbens (NAc) following chronic genetic or pharmacological DBH inhibition, and overexpression of ßArr2 in the NAc normalized cocaine-induced locomotion in DBH knockout (Dbh -/-) mice. The D2/3 agonist quinpirole decreased excitability in NAc medium spiny neurons (MSNs) from control, but not Dbh -/- animals, where instead there was a trend for an excitatory effect. The Gαi inhibitor NF023 abolished the quinpirole-induced decrease in excitability in control MSNs, but had no effect in Dbh -/- MSNs, whereas the Gαs inhibitor NF449 restored the ability of quinpirole to decrease excitability in Dbh -/- MSNs, but had no effect in control MSNs. These results suggest that chronic loss of noradrenergic tone alters behavioral responses to cocaine via decreases in ßArr2 and cellular responses to D2/D3 activation, potentially via changes in D2-like receptor G-protein coupling in NAc MSNs.

Transcriptional and translational heterogeneity among neonatal mouse spermatogonia.

Spermatogonial stem cells (SSCs) are a subset of undifferentiated spermatogonia responsible for ongoing spermatogenesis in mammalian testes. Spermatogonial stem cells arise from morphologically homogeneous prospermatogonia, but growing evidence suggests that only a subset of prospermatogonia develops into the foundational SSC pool. This predicts that subtypes of undifferentiated spermatogonia with discrete mRNA and protein signatures should be distinguishable in neonatal testes. We used single-cell quantitative RT-PCR to examine mRNA levels of 172 genes in individual spermatogonia from 6-day postnatal (P6) mouse testes. Cells enriched from P6 testes using the StaPut or THY1(+) magnetic cell sorting methods exhibited considerable heterogeneity in the abundance of specific germ cell and stem cell mRNAs, segregating into one somatic and three distinct spermatogonial clusters. However, P6 Id4-eGFP(+) transgenic spermatogonia, which are known to be enriched for SSCs, were more homogeneous in their mRNA levels, exhibiting uniform levels for the majority of genes examined (122 of 172). Interestingly, these cells displayed nonuniform (50 of 172) expression of a smaller cohort of these genes, suggesting there is substantial heterogeneity even within the Id4-eGFP(+) population. Further, although immunofluorescence staining largely demonstrated conformity between mRNA and protein levels, some proteins were observed in patterns that were disparate from those detected for the corresponding mRNAs in Id4-eGFP(+) spermatogonia (e.g., Kit, Sohlh2, Stra8), suggesting additional heterogeneity is introduced at the posttranscriptional level. Taken together, these data demonstrate the existence of multiple spermatogonial subtypes in P6 mouse testes and raise the intriguing possibility that these subpopulations may correlate with the development of functionally distinct spermatogenic cell types.

Food restriction increases glutamate receptor-mediated burst firing of dopamine neurons.

Restriction of food intake increases the acquisition of drug abuse behavior and enhances the reinforcing efficacy of those drugs. However, the neurophysiological mechanisms responsible for the interactions between feeding state and drug use are largely unknown. Here we show that chronic mild food restriction increases the burst firing of dopamine neurons in the substantia nigra. Dopamine neurons from food-restricted mice exhibited increased burst firing in vivo, an effect that was enhanced by an injection of the psychomotor stimulant cocaine (10 mg/kg, i.p.). Food restriction also enhanced aspartic acid-induced burst firing of dopamine neurons in an ex vivo brain slice preparation, consistent with an adaptation occurring in the somatodendritic compartment and independent of a circuit mechanism. Enhanced burst firing persisted after 10 d of free feeding following chronic food restriction but was not observed following a single overnight fast. Whole-cell patch-clamp recordings indicated that food restriction also increased electrically evoked AMPAR/NMDAR ratios and increased D2 autoreceptor-mediated desensitization in dopamine neurons. These results identify dopamine neurons in the substantia nigra as a convergence point for the interactions between feeding state and drugs of abuse. Furthermore, increased glutamate transmission combined with decreased autoreceptor inhibition could work in concert to enhance drug efficacy in response to food restriction.

Before It Gets Started: Regulating Translation at the 5' UTR.

Translation regulation plays important roles in both normal physiological conditions and diseases states. This regulation requires cis-regulatory elements located mostly in 5' and 3' UTRs and trans-regulatory factors (e.g., RNA binding proteins (RBPs)) which recognize specific RNA features and interact with the translation machinery to modulate its activity. In this paper, we discuss important aspects of 5' UTR-mediated regulation by providing an overview of the characteristics and the function of the main elements present in this region, like uORF (upstream open reading frame), secondary structures, and RBPs binding motifs and different mechanisms of translation regulation and the impact they have on gene expression and human health when deregulated.

Enriching for correct prediction of biological processes using a combination of diverse classifiers.

BACKGROUND: Machine learning models (classifiers) for classifying genes to biological processes each have their own unique characteristics in what genes can be classified and to what biological processes. No single learning model is qualitatively superior to any other model and overall precision for each model tends to be low. The classification results for each classifier can be complementary and synergistic suggesting the benefit of a combination of algorithms, but often the prediction probability outputs of various learning models are neither comparable nor compatible for combining. A means to compare outputs regardless of the model and data used and combine the results into an improved comprehensive model is needed. RESULTS: Gene expression patterns from NCI's panel of 60 cell lines were used to train a Random Forest, a Support Vector Machine and a Neural Network model, plus two over-sampled models for classifying genes to biological processes. Each model produced unique characteristics in the classification results. We introduce the Precision Index measure (PIN) from the maximum posterior probability that allows assessing, comparing and combining multiple classifiers. The class specific precision measure (PIC) is introduced and used to select a subset of predictions across all classes and all classifiers with high precision. We developed a single classifier that combines the PINs from these five models in prediction and found that the PIN Combined Classifier (PINCom) significantly increased the number of correctly predicted genes over any single classifier. The PINCom applied to test genes that were not used in training also showed substantial improvement over any single model. CONCLUSIONS: This paper introduces novel and effective ways of assessing predictions by their precision and recall plus a method that combines several machine learning models and capitalizes on synergy and complementation in class selection, resulting in higher precision and recall. Different machine learning models yielded incongruent results each of which were successfully combined into one superior model using the PIN measure we developed. Validation of the boosted predictions for gene functions showed the genes to be accurately predicted.

Sequence signatures and mRNA concentration can explain two-thirds of protein abundance variation in a human cell line.

Transcription, mRNA decay, translation and protein degradation are essential processes during eukaryotic gene expression, but their relative global contributions to steady-state protein concentrations in multi-cellular eukaryotes are largely unknown. Using measurements of absolute protein and mRNA abundances in cellular lysate from the human Daoy medulloblastoma cell line, we quantitatively evaluate the impact of mRNA concentration and sequence features implicated in translation and protein degradation on protein expression. Sequence features related to translation and protein degradation have an impact similar to that of mRNA abundance, and their combined contribution explains two-thirds of protein abundance variation. mRNA sequence lengths, amino-acid properties, upstream open reading frames and secondary structures in the 5' untranslated region (UTR) were the strongest individual correlates of protein concentrations. In a combined model, characteristics of the coding region and the 3'UTR explained a larger proportion of protein abundance variation than characteristics of the 5'UTR. The absolute protein and mRNA concentration measurements for >1000 human genes described here represent one of the largest datasets currently available, and reveal both general trends and specific examples of post-transcriptional regulation.

The effect of gonadectomy and estradiol on sensitivity to oxidative stress.

The sexual dimorphism of life span and caloric restriction effects in numerous species suggest that estradiol (E2) is protective against oxidative damage. The only direct test of E2's protective effect in mice against in vivo oxidative stress to date may have been confounded by E2's direct chemical action as an antioxidant because it was administered at very high dosages. Therefore, we have identified a low yet physiologically effective dose of E2. We then administered this dose using subcutaneous time-release pellets to ovariectomized mice. Two weeks after E2 pellet implantation, sham-operated, ovariectomized, and ovariectomized E2-supplemented female mice were injected with a lethal dose of paraquat and their survival was followed. It was observed that ovariectomy exacerbates paraquat-induced mortality and is rescued by E2 supplementation. An equivalent experiment was performed on sham-operated, orchidectomized, and E2-supplemented orchidectomized male mice. The survival of male mice was improved by orchidectomy, and E2 gave no further benefit. We interpret the results to mean that E2 is protective against oxidative stress through its regulatory role and that testosterone diminishes protection against oxidative stress.

Genomic analyses of musashi1 downstream targets show a strong association with cancer-related processes.

Musashi1 (Msi1) is a highly conserved RNA-binding protein with pivotal functions in stem cell maintenance, nervous system development, and tumorigenesis. Despite its importance, only three direct mRNA targets have been characterized so far: m-numb, CDKN1A, and c-mos. Msi1 has been shown to affect their translation by binding to short elements located in the 3'-untranslated region. To better understand Msi1 functions, we initially performed an RIP-Chip analysis in HEK293T cells; this method consists of isolation of specific RNA-protein complexes followed by identification of the RNA component via microarrays. A group of 64 mRNAs was found to be enriched in the Msi1-associated population compared with controls. These genes belong to two main functional categories pertinent to tumorigenesis: 1) cell cycle, cell proliferation, cell differentiation, and apoptosis and 2) protein modification (including ubiquitination and ubiquitin cycle). To corroborate our findings, we examined the impact of Msi1 expression on both mRNA (transcriptomic) and protein (proteomic) expression levels. Genes whose mRNA levels were affected by Msi1 expression have a Gene Ontology distribution similar to RIP-Chip results, reinforcing Msi1 participation in cancer-related processes. The proteomics study revealed that Msi1 can have either positive or negative effects on gene expression of its direct targets. In summary, our results indicate that Msi1 affects a network of genes and could function as a master regulator during development and tumor formation.

Over-represented sequences located on 3' UTRs are potentially involved in regulatory functions.

Eukaryotic gene expression must be coordinated for the proper functioning of biological processes. This coordination can be achieved both at the transcriptional and post-transcriptional levels. In both cases, regulatory sequences placed at either promoter regions or on UTRs function as markers recognized by regulators that can then activate or repress different groups of genes according to necessity. While regulatory sequences involved in transcription are quite well documented, there is a lack of information on sequence elements involved in post-transcriptional regulation. We used a statistical over-representation method to identify novel regulatory elements located on UTRs. An exhaustive search approach was used to calculate the frequency of all possible n-mers (short nucleotide sequences) in 16,160 human genes of NCBI RefSeq sequences and to identify any peculiar usage of n-mers on UTRs. After a stringent filtering process, we identified 2,772 highly over-represented n-mers on 3' UTRs. We provide evidence that these n-mers are potentially involved in regulatory functions. Identified n-mers overlap with previously identified binding sites for HuR and TIA-1 and, ARE and GRE sequences. We determine also that n-mers overlap with predicted miRNA target sites. Finally, a method to cluster n-mer groups allowed the identification of putative gene networks.

The origin of chromosome imbalances in neuroblastoma.

Many recurrent large-scale chromosome abnormalities associated with poor clinical outcomes have been identified in neuroblastoma, a pediatric tumor that accounts for 15% of childhood cancer deaths. We have previously used high-resolution oligonucleotide array comparative genomic hybridization to map 461 chromosome breakpoints leading to large-scale chromosome imbalances in 56 primary neuroblastoma tumors and cell lines. Here, we analyze the distribution of DNA sequence elements and genomic landmarks found within these breakpoint intervals and in 15,800 randomly generated intervals of similar size. The most consistent finding was that neuroblastoma chromosome breakpoints occur preferentially in GC-rich regions of the genome. It is not unsurprising that these regions have fewer (AT)(n) microsatellite repeat sequences. In addition, chromosome breakpoints occurring in neuroblastoma also appeared to be preferentially associated with ancestral chromosome breakpoint regions on several chromosomes, suggesting that such sites also act as hotspots for chromosome rearrangement in somatic cells. Very little evidence for the enrichment of Alu and other types of repeats in breakpoint intervals was obtained. Overall, our results are consistent with a mechanistic model involving nonhomologous end joining of DNA double-strand breaks that have been generated in a nonrandom manner.

Gene function classification using NCI-60 cell line gene expression profiles.

Gene expression patterns from NCI's panel of 60 cell lines were used to train a Neural Network model for classifying genes to pathways. The model assigns probabilities to each gene for each of the 21 modeled pathways assigned by the Kyoto Encyclopedia of Genes and Genomes. Cross-validation of the model showed that 10 of the 21 pathways exhibited good performance in statistical significance and accuracy. The model was designed to output gene probabilities that could be screened for higher probabilities resulting in higher confidence in classification though yielding fewer genes per pathway. The model was deployed on 5798 genes and our approach allowed us to ascertain the most relevant genes above an estimated background. Eight pathways were identified with both good cross-validation and significant numbers above background, TCA Cycle, Oxidative Phosphorylation, Porphyrin Biosynthesis, Ribosome, Polymerases, Proteasome, Cell Cycle, and Cell Adhesion. Gene Ontology (GO) annotation was used for additional validation of gene classification results. A total of 551 GO annotated genes and 468 unannotated genes were classified to the 8 pathways. The primary and secondary classifications of genes revealed known pathway relationships and provide the potential for discovering new pathway relationships.