An Efficient, Parallelized Algorithm for Optimal Conditional Entropy-Based Feature Selection.

In Machine Learning, feature selection is an important step in classifier design. It consists of finding a subset of features that is optimum for a given cost function. One possibility to solve feature selection is to organize all possible feature subsets into a Boolean lattice and to exploit the fact that the costs of chains in that lattice describe U-shaped curves. Minimization of such cost function is known as the U-curve problem. Recently, a study proposed U-Curve Search (UCS), an optimal algorithm for that problem, which was successfully used for feature selection. However, despite of the algorithm optimality, the UCS required time in computational assays was exponential on the number of features. Here, we report that such scalability issue arises due to the fact that the U-curve problem is NP-hard. In the sequence, we introduce the Parallel U-Curve Search (PUCS), a new algorithm for the U-curve problem. In PUCS, we present a novel way to partition the search space into smaller Boolean lattices, thus rendering the algorithm highly parallelizable. We also provide computational assays with both synthetic data and Machine Learning datasets, where the PUCS performance was assessed against UCS and other golden standard algorithms in feature selection.

Proteomic and Glycoproteomic Profilings Reveal That Post-translational Modifications of Toxins Contribute to Venom Phenotype in Snakes.

Snake venoms are biological weapon systems composed of secreted proteins and peptides that are used for immobilizing or killing prey. Although post-translational modifications are widely investigated because of their importance in many biological phenomena, we currently still have little understanding of how protein glycosylation impacts the variation and stability of venom proteomes. To address these issues, here we characterized the venom proteomes of seven Bothrops snakes using a shotgun proteomics strategy. Moreover, we compared the electrophoretic profiles of native and deglycosylated venoms and, in order to assess their subproteomes of glycoproteins, we identified the proteins with affinity for three lectins with different saccharide specificities and their putative glycosylation sites. As proteinases are abundant glycosylated toxins, we examined the effect of N-deglycosylation on their catalytic activities and show that the proteinases of the seven venoms were similarly affected by removal of N-glycans. Moreover, we prospected putative glycosylation sites of transcripts of a B. jararaca venom gland data set and detected toxin family related patterns of glycosylation. Based on our global analysis, we report that Bothrops venom proteomes and glycoproteomes contain a core of components that markedly define their composition, which is conserved upon evolution in parallel to other molecular markers that determine their phylogenetic classification.

Autophagy buffers Ras-induced genotoxic stress enabling malignant transformation in keratinocytes primed by human papillomavirus.

Malignant transformation involves an orchestrated rearrangement of cell cycle regulation mechanisms that must balance autonomic mitogenic impulses and deleterious oncogenic stress. Human papillomavirus (HPV) infection is highly prevalent in populations around the globe, whereas the incidence of cervical cancer is 0.15%. Since HPV infection primes cervical keratinocytes to undergo malignant transformation, we can assume that the balance between transforming mitogenic signals and oncogenic stress is rarely attained. We showed that highly transforming mitogenic signals triggered by HRasG12V activity in E6E7-HPV-keratinocytes generate strong replication and oxidative stresses. These stresses are counteracted by autophagy induction that buffers the rapid increase of ROS that is the main cause of genotoxic stress promoted by the oncoprotein. As a result, autophagy creates a narrow window of opportunity for malignant keratinocytes to emerge. This work shows that autophagy is crucial to allow the transition of E6E7 keratinocytes from an immortalized to a malignant state caused by HRasG12V.

Comparative analysis of the high molecular mass subproteomes of eight Bothrops snake venoms.

Snake venoms are extremely active biological secretions composed primarily of various classes of enzymes. The genus Bothrops comprises various pit viper species that represent the most medically significant taxa in Central and South America, accounting for more human envenomations and fatalities than any other snakes in the region. Venom proteomes of many Bothrops species have been well-characterized but investigations have focused almost exclusively on proteins smaller than 100 kDa despite expression of larger components being documented in several Bothrops venoms. This study sought to achieve detailed identification of major components in the high molecular mass subproteome of venoms from eight Bothrops species (B. brazili, B. cotiara, B. insularis, B. jararaca, B. jararacussu, B. leucurus, B. moojeni and B. neuwiedi). Enzymes such as metalloproteinases and L-amino acid oxidases were the most prominent components identified in the first size-exclusion chromatography fractions of these venoms. Minor components also identified in the first peaks included 5'-nucleotidase, aminopeptidase, phosphodiesterase, and phospholipases A2 and B. Most of these components disappeared in electrophoretic profiles under reducing conditions, suggesting that they may be composed of more than one polypeptide chain. A significant shift in the molecular masses of these protein bands was observed following enzymatic N-deglycosylation, indicating that they may contain N-glycans. Furthermore, none of the identified high molecular mass proteins were shared by all eight species, revealing a high level of interspecific variability among these venom components.

Transcription activity contributes to the firing of non-constitutive origins in African trypanosomes helping to maintain robustness in S-phase duration.

The co-synthesis of DNA and RNA potentially generates conflicts between replication and transcription, which can lead to genomic instability. In trypanosomatids, eukaryotic parasites that perform polycistronic transcription, this phenomenon and its consequences are still little studied. Here, we showed that the number of constitutive origins mapped in the Trypanosoma brucei genome is less than the minimum required to complete replication within S-phase duration. By the development of a mechanistic model of DNA replication considering replication-transcription conflicts and using immunofluorescence assays and DNA combing approaches, we demonstrated that the activation of non-constitutive (backup) origins are indispensable for replication to be completed within S-phase period. Together, our findings suggest that transcription activity during S phase generates R-loops, which contributes to the emergence of DNA lesions, leading to the firing of backup origins that help maintain robustness in S-phase duration. The usage of this increased pool of origins, contributing to the maintenance of DNA replication, seems to be of paramount importance for the survival of this parasite that affects million people around the world.

Fibroblast Growth Factor 2 lethally sensitizes cancer cells to stress-targeted therapeutic inhibitors.

In malignant transformation, cellular stress-response pathways are dynamically mobilized to counterbalance oncogenic activity, keeping cancer cells viable. Therapeutic disruption of this vulnerable homeostasis might change the outcome of many human cancers, particularly those for which no effective therapy is available. Here, we report the use of fibroblast growth factor 2 (FGF2) to demonstrate that further mitogenic activation disrupts cellular homeostasis and strongly sensitizes cancer cells to stress-targeted therapeutic inhibitors. We show that FGF2 enhanced replication and proteotoxic stresses in a K-Ras-driven murine cancer cell model, and combinations of FGF2 and proteasome or DNA damage response-checkpoint inhibitors triggered cell death. CRISPR/Cas9-mediated K-Ras depletion suppressed the malignant phenotype and prevented these synergic toxicities in these murine cells. Moreover, in a panel of human Ewing's sarcoma family tumor cells, sublethal concentrations of bortezomib (proteasome inhibitor) or VE-821 (ATR inhibitor) induced cell death when combined with FGF2. Sustained MAPK-ERK1/2 overactivation induced by FGF2 appears to underlie these synthetic lethalities, as late pharmacological inhibition of this pathway restored cell homeostasis and prevented these described synergies. Our results highlight how mitotic signaling pathways which are frequently overridden in malignant transformation might be exploited to disrupt the robustness of cancer cells, ultimately sensitizing them to stress-targeted therapies. This approach provides a new therapeutic rationale for human cancers, with important implications for tumors still lacking effective treatment, and for those that frequently relapse after treatment with available therapies.

An Interdisciplinary Approach for Designing Kinetic Models of the Ras/MAPK Signaling Pathway.

We present in this article a methodology for designing kinetic models of molecular signaling networks, which was exemplarily applied for modeling one of the Ras/MAPK signaling pathways in the mouse Y1 adrenocortical cell line. The methodology is interdisciplinary, that is, it was developed in a way that both dry and wet lab teams worked together along the whole modeling process.

A New Approach for Identification of Cancer-related Pathways using Protein Networks and Genomic Data.

Cancer cells have anomalous development and proliferation due to disturbances in their control systems. The study of the behavior of cellular control system requires high-throughput dynamical data. Unfortunately, this type of data is not largely available. This fact motivates the main issue of this article: how to use static omics data and available biological knowledge to get new information about the elements of the control system in cancer cells. Two important measures to access the state of the cellular control system are the gene expression profile and the signaling pathways. This article uses a combination of these two static omics data to gain insights on the states of a cancer cell. To extract information from this kind of data, a statistical computational model was formalized and implemented. In order to exemplify the application of some aspects of the developed conceptual framework, we verified the hypothesis that different types of cancer cells have different disturbed signaling pathways. To this end, we developed a method that recovers small protein networks, called motifs, which are differentially represented in some subtypes of breast cancer. These differentially represented motifs are enriched with specific gene ontologies as well as with new putative cancer genes.

Bioinfomatics for the Citrus EST Project (CitEST)

In this work we describe all the computational environments, pipelines, and web services developed for the CitEST transcriptome project, on which all the annotation researchers relied. We also present a complete list of CitEST libraries and, for each of them, the general features after the in silico processing, showing some quantitative information.

In silico prediction of gene expression patterns in Citrus flavedo

Out of the 18,942 flavedo expressed sequences (clusters plus singletons) in Citrus sinensis from the Citrus EST Project (CitEST), 25 were statistically supported to be differentially expressed in this tissue after a double in silico hybridization strategy against leaf-, flower-, and bark-derived ESTs. Five of them, two terpene synthases and three O-methyltransferases, are absent in the other citrus tissues with concomitant 2x2 statistics, supporting the hypothesis that they are putative flavedo-specific expressed sequences. The pattern of these differentially expressed sequences during fruit development suggests that most of them are developmentally regulated. Some expressed gene products, including a putative germin-like protein highly expressed in flavedo, are shown to be promising candidates for further characterization. In addition to promoter seeking, this kind of analysis can lead to gene discovery, tissue-specific and tissue-enriched expression pattern predictions (as shown herein) and can also be adopted as an in silico first, and probably reliable approach, for detecting expression profiles from EST sequencing efforts before experimental validation is available or for heuristically guiding that validation.

In silico analysis of ESTs from roots of Rangpur lime (Citrus limonia Osbeck) under water stress

CitEST project resulted in the construction of cDNA libraries from different Citrus sp. tissues under various physiological conditions. Among them, plantlets of Rangpur lime were exposed to hydroponic conditions with and without water stress using PEG6000. RNA from roots was obtained and generated a total of 4,130 valid cDNA reads, with 2,020 from the non-stressed condition and 2,110 from the stressed set. Bioinformatic analyses measured the frequency of each read in the libraries and yielded an in silico transcriptional profile for each condition. A total of 40 contigs were differentially expressed and allowed to detect up-regulated homologue sequences to well known genes involved in stress response, such as aquaporins, dehydrin, sucrose synthase, and proline-related synthase. Some sequences could not be classified by using FunCat and remained with an unknown function. A large number of sequences presented high similarities to annotated genes involved with cell energy, protein synthesis and cellular transport, suggesting that Rangpur lime may sustain active cell growth under stressed condition. The presence of membrane transporters and cell signaling components could be an indication of a coordinated morphological adaptation and biochemical response during drought, helping to explain the higher tolerance of this rootstock to water stress.

Differential expression of genes identified from Poncirus trifoliata tissue inoculated with CTV through EST analysis and in silico hybridization

Citrus is the most important fruit crop in Brazil and Citrus tristeza virus (CTV) is considered one of the most important pathogens of citrus. Most citrus species and varieties are susceptible to CTV infection. However, Poncirus trifoliata, a close relative of citrus, is resistant to the virus. In order to better understand the responses of citrus plants to the infection of CTV, we constructed expressed sequence tag (EST) libraries with tissues collected from Poncirus trifoliata plants, inoculated or not with Citrus tristeza virus at 90 days after inoculation, grafted on Rangpur lime rootstocks. We generated 17,867 sequence tags from Poncirus trifoliata inoculated (8,926 reads) and not (8,941 reads) with a severe CTV isolate. A total of 2,782 TCs (Tentative Consensi sequences) were obtained using both cDNA libraries in a single clusterization procedure. By the in silico hybridization approach, 289 TCs were identified as differentially expressed in the two libraries. A total of 121 TCs were found to be overexpressed in plants infected with CTV and were grouped in 12 primary functional categories. The majority of them were associated with metabolism and defense response. Some others were related to lignin, ethylene biosynthesis and PR proteins. In general, the differentially expressed transcripts seem to be somehow involved in secondary plant response to CTV infection.

Analysis of expressed sequence tags from Citrus sinensis L. Osbeck infected with Xylella fastidiosa

In order to understand the genetic responses resulting from physiological changes that occur in plants displaying citrus variegated chlorosis (CVC) symptoms, we adopted a strategy of comparing two EST libraries from sweet orange [Citrus sinensis (L.) Osbeck]. One of them was prepared with plants showing typical CVC symptoms caused by Xylella fastidiosa and the other with non-inoculated plants. We obtained 15,944 ESTs by sequencing the two cDNA libraries. Using an in silico hybridization strategy, 37 genes were found to have significant variation at the transcriptional level. Within this subset, 21 were up-regulated and 16 were down-regulated in plants with CVC. The main functional categories of the down-regulated transcripts in plants with CVC were associated with metabolism, protein modification, energy and transport facilitation. The majority of the up-regulated transcripts were associated with metabolism and defense response. Some transcripts associated with adaptation to stress conditions were up-regulated in plants with CVC and could explain why plants remain alive even under severe water and nutritional stress. Others of the up-regulated transcripts are related to defense response suggesting that sweet orange plants activate their defense machinery. The genes associated with stress response might be expressed as part of a secondary response related to physiological alterations caused by the infection.