Structural Parameters and Behavior in Simulated Body Fluid of High Entropy Alloy Thin Films.

The structure, composition and corrosion properties of thin films synthesized using the Pulsed Laser Deposition (PLD) technique starting from a three high entropy alloy (HEA) AlCoCrFeNix produced by vacuum arc remelting (VAR) method were investigated. The depositions were performed at room temperature on Si and mirror-like polished Ti substrates either under residual vacuum (low 10-7 mbar, films denoted HEA2, HEA6, and HEA10, which were grown from targets with Ni concentration molar ratio, x, equal to 0.4, 1.2, and 2.0, respectively) or under N2 (10-4 mbar, films denoted HEN2, HEN6, and HEN10 for the same Ni concentration molar ratios). The deposited films' structures, investigated using Grazing Incidence X-ray Diffraction, showed the presence of face-centered cubic and body-centered cubic phases, while their surface morphology, investigated using scanning electron microscopy, exhibited a smooth surface with micrometer size droplets. The mass density and thickness were obtained from simulations of acquired X-ray reflectivity curves. The films' elemental composition, estimated using the energy dispersion X-ray spectroscopy, was quite close to that of the targets used. X-ray Photoelectron Spectroscopy investigation showed that films deposited under a N2 atmosphere contained several percentages of N atoms in metallic nitride compounds. The electrochemical behavior of films under simulated body fluid (SBF) conditions was investigated by Open Circuit Potential (OCP) and Electrochemical Impedance Spectroscopy measurements. The measured OCP values increased over time, implying that a passive layer was formed on the surface of the films. It was observed that all films started to passivate in SBF solution, with the HEN6 film exhibiting the highest increase. The highest repassivation potential was exhibited by the same film, implying that it had the highest stability range of all analyzed films. Impedance measurements indicated high corrosion resistance values for HEA2, HEA6, and HEN6 samples. Much lower resistances were found for HEN10 and HEN2. Overall, HEN6 films exhibited the best corrosion behavior among the investigated films. It was noticed that for 24 h of immersion in SBF solution, this film was also a physical barrier to the corrosion process, not only a chemical one.

Identifying transcriptomic correlates of histology using deep learning.

Linking phenotypes to specific gene expression profiles is an extremely important problem in biology, which has been approached mainly by correlation methods or, more fundamentally, by studying the effects of gene perturbations. However, genome-wide perturbations involve extensive experimental efforts, which may be prohibitive for certain organisms. On the other hand, the characterization of the various phenotypes frequently requires an expert's subjective interpretation, such as a histopathologist's description of tissue slide images in terms of complex visual features (e.g. 'acinar structures'). In this paper, we use Deep Learning to eliminate the inherent subjective nature of these visual histological features and link them to genomic data, thus establishing a more precisely quantifiable correlation between transcriptomes and phenotypes. Using a dataset of whole slide images with matching gene expression data from 39 normal tissue types, we first developed a Deep Learning tissue classifier with an accuracy of 94%. Then we searched for genes whose expression correlates with features inferred by the classifier and demonstrate that Deep Learning can automatically derive visual (phenotypical) features that are well correlated with the transcriptome and therefore biologically interpretable. As we are particularly concerned with interpretability and explainability of the inferred histological models, we also develop visualizations of the inferred features and compare them with gene expression patterns determined by immunohistochemistry. This can be viewed as a first step toward bridging the gap between the level of genes and the cellular organization of tissues.

Correction: Exploring the reproducibility of functional connectivity alterations in Parkinson's disease.

[This corrects the article DOI: 10.1371/journal.pone.0188196.].

Exploring the reproducibility of functional connectivity alterations in Parkinson's disease.

Since anatomic MRI is presently not able to directly discern neuronal loss in Parkinson's Disease (PD), studying the associated functional connectivity (FC) changes seems a promising approach toward developing non-invasive and non-radioactive neuroimaging markers for this disease. While several groups have reported such FC changes in PD, there are also significant discrepancies between studies. Investigating the reproducibility of PD-related FC changes on independent datasets is therefore of crucial importance. We acquired resting-state fMRI scans for 43 subjects (27 patients and 16 normal controls, with 2 replicate scans per subject) and compared the observed FC changes with those obtained in two independent datasets, one made available by the PPMI consortium (91 patients, 18 controls) and a second one by the group of Tao Wu (20 patients, 20 controls). Unfortunately, PD-related functional connectivity changes turned out to be non-reproducible across datasets. This could be due to disease heterogeneity, but also to technical differences. To distinguish between the two, we devised a method to directly check for disease heterogeneity using random splits of a single dataset. Since we still observe non-reproducibility in a large fraction of random splits of the same dataset, we conclude that functional heterogeneity may be a dominating factor behind the lack of reproducibility of FC alterations in different rs-fMRI studies of PD. While global PD-related functional connectivity changes were non-reproducible across datasets, we identified a few individual brain region pairs with marginally consistent FC changes across all three datasets. However, training classifiers on each one of the three datasets to discriminate PD scans from controls produced only low accuracies on the remaining two test datasets. Moreover, classifiers trained and tested on random splits of the same dataset (which are technically homogeneous) also had low test accuracies, directly substantiating disease heterogeneity.

Increased connectivity between sensorimotor and attentional areas in Parkinson's disease.

INTRODUCTION: Our study is using Independent Component Analysis (ICA) to evaluate functional connectivity changes in Parkinson's disease (PD) in an unbiased manner. METHODS: Resting-state functional magnetic resonance imaging (rs-fMRI) data was collected for 27 PD patients and 16 healthy subjects. Differences for intra- and inter-network connectivity between healthy subjects and patients were investigated using FMRIB Software Library (FSL) tools (Melodic ICA, dual regression, FSLNets). RESULTS: Twenty-three ICA maps were identified as components of neuronal origin. For intra-network connectivity changes, eight components showed a significant connectivity increase in patients (p < 0.05); these were correlated with clinical scores and were largest for (sensori)motor networks. For inter-network connectivity changes, we found higher connectivity between the sensorimotor network and the spatial attention network (p = 0.0098) and lower connectivity between anterior and posterior default mode networks (DMN) (p = 0.024), anterior DMN and visual recognition networks (p = 0.026), as well as between visual attention and main dorsal attention networks (p = 0.03), for patients as compared to healthy subjects. The area under the Receiver Operating Characteristics (ROC) curve for the best predictor (partial correlation between sensorimotor and spatial attention networks) was 0.772. These functional alterations were not associated with any gray or white matter structural changes. CONCLUSION: Our results show higher connectivity between sensorimotor and spatial attention areas in patients that may be related to the reduced movement automaticity in PD.

Extracting significant sample-specific cancer mutations using their protein interactions.

We present a joint analysis method for mutation and gene expression data employing information about proteins that are highly interconnected at the level of protein to protein (pp) interactions, which we apply to the TCGA Acute Myeloid Leukemia (AML) dataset. Given the low incidence of most mutations in virtually all cancer types, as well as the significant inter-patient heterogeneity of the mutation landscape, determining the true causal mutations in each individual patient remains one of the most important challenges for personalized cancer diagnostics and therapy. More automated methods are needed for determining these "driver" mutations in each individual patient. For this purpose, we are exploiting two types of contextual information: (1) the pp interactions of the mutated genes, as well as (2) their potential correlations with gene expression clusters. The use of pp interactions is based on our surprising finding that most AML mutations tend to affect nontrivial protein to protein interaction cliques.

Assessment of the stromal contribution to Sonic Hedgehog-dependent pancreatic adenocarcinoma.

Pancreatic ductal adenocarcinoma (PDAC) remains one of the most lethal malignancies. It is typically detected at an advanced stage, at which the therapeutic options are very limited. One remarkable feature of PDAC that contributes to its resilience to treatment is the extreme stromal activation seen in these tumors. Often, the vast majority of tumor bulk consists of non-tumor cells that together provide a tumor-promoting environment. One of the signals that maintains and activates the stroma is the developmental protein Sonic Hedgehog (SHH). As the disease progresses, tumor cells produce increasing amounts of SHH, which activates the surrounding stroma to aid in tumor progression. To better understand this response and identify targets for inhibition, we aimed to elucidate the proteins that mediate the SHH-driven stromal response in PDAC. For this a novel mixed-species coculture model was set up in which the cancer cells are human, and the stroma is modeled by mouse fibroblasts. In conjunction with next-generation sequencing we were able to use the sequence difference between these species to genetically distinguish between the epithelial and stromal responses to SHH. The stromal SHH-dependent genes from this analysis were validated and their relevance for human disease was subsequently determined in two independent patient cohorts. In non-microdissected tissue from PDAC patients, in which a large amount of stroma is present, the targets were confirmed to associate with tumor stroma versus normal pancreatic tissue. Patient survival analysis and immunohistochemistry identified CDA, EDIL3, ITGB4, PLAUR and SPOCK1 as SHH-dependent stromal factors that are associated with poor prognosis in PDAC patients. Summarizing, the presented data provide insight into the role of the activated stroma in PDAC, and how SHH acts to mediate this response. In addition, the study has yielded several candidates that are interesting therapeutic targets for a disease for which treatment options are still inadequate.

Identification of potential biomarkers for early and advanced gastric adenocarcinoma detection.

BACKGROUND/AIMS: This study aimed to understand gradual biological variations during gastric tumorigenesis, and to identify the candidate genes that are involved in tumor progression and metastasis. METHODOLOGY: cDNA microarray data were obtained from 10 pair of cancerous and normal adjacent tissue from gastric adenocarcinoma patients. The samples were divided in primary and advanced gastric adenocarcinoma with lymph node metastasis. Validation of the microarray data was accomplished by quantitative RT-PCR on additional 41 samples. The significantly modified genes were grouped in clusters according to their functional annotation, and comparison was done regarding molecular mechanisms involved tumor progression. RESULTS: A total of 136 genes were up-regulated and 96 genes were down-regulated by at least fourfold in tumor tissue. The analysis of gene clusters revealed a complex remodelling of normal gastric epithelium morphology and function associated with the tumorigenesis and metastasis. A large number of proteases are being overexpressed, together with keratins, genes associated with morphogenesis and anti-apoptosis. Between the most significant down-regulated genes, there were genes involved in gastric motility and synthesis and genes related to metabolic and pro-apoptotic processes. We also report, the identification of seven genes, significant up-regulated, that seem to be associated with tumor progression: KRT17, COL10A2, KIAA1199, SPP1, IL11, S100A2, and MMP3. CONCLUSIONS: Our cDNA microarray study identified several genes that appeared to meet the criteria of a good biomarker, and may therefore be especially useful for the development of diagnostic tools, for the early detection, or for the prediction of tumor progression.

Extracting gene expression profiles common to colon and pancreatic adenocarcinoma using simultaneous nonnegative matrix factorization.

In this paper we introduce a clustering algorithm capable of simultaneously factorizing two distinct gene expression datasets with the aim of uncovering gene regulatory programs that are common to the two phenotypes. The siNMF algorithm simultaneously searches for two factorizations that share the same gene expression profiles. The two key ingredients of this algorithm are the nonnegativity constraint and the offset variables, which together ensure the sparseness of the factorizations. While cancer is a very heterogeneous disease, there is overwhelming recent evidence that the differences between cancer subtypes implicate entire pathways and biological processes involving large numbers of genes, rather than changes in single genes. We have applied our simultaneous factorization algorithm looking for gene expression profiles that are common between the more homogeneous pancreatic ductal adenocarcinoma (PDAC) and the more heterogeneous colon adenocarcinoma. The fact that the PDAC signature is active in a large fraction of colon adeocarcinoma suggests that the oncogenic mechanisms involved may be similar to those in PDAC, at least in this subset of colon samples. There are many approaches to uncovering common mechanisms involved in different phenotypes, but most are based on comparing gene lists. The approach presented in this paper additionally takes gene expression data into account and can thus be more sensitive.

Combined gene expression analysis of whole-tissue and microdissected pancreatic ductal adenocarcinoma identifies genes specifically overexpressed in tumor epithelia.

BACKGROUND/AIMS: The precise details of pancreatic ductal adenocarcinoma (PDAC) pathogenesis are still insufficiently known, requiring the use of high-throughput methods. However, PDAC is especially difficult to study using microarrays due to its strong desmoplastic reaction, which involves a hyperproliferating stroma that effectively "masks" the contribution of the minoritary neoplastic epithelial cells. Thus it is not clear which of the genes that have been found differentially expressed between normal and whole tumor tissues are due to the tumor epithelia and which simply reflect the differences in cellular composition. To address this problem, laser microdissection studies have been performed, but these have to deal with much smaller tissue sample quantities and therefore have significantly higher experimental noise. METHODOLOGY: In this paper we combine our own large sample whole-tissue study with a previously published smaller sample microdissection study by Grützmann et al. to identify the genes that are specifically overexpressed in PDAC tumor epithelia. RESULTS: The overlap of this list of genes with other microarray studies of pancreatic cancer as well as with the published literature is impressive. Moreover, we find a number of genes whose over-expression appears to be inversely correlated with patient survival: keratin 7, laminin gamma 2, stratifin, platelet phosphofructokinase, annexin A2, MAP4K4 and OACT2 (MBOAT2), which are all specifically upregulated in the neoplastic epithelia, rather than the tumor stroma. CONCLUSIONS: We improve on other microarray studies of PDAC by putting together the higher statistical power due to a larger number of samples with information about cell-type specific expression and patient survival.

Sparse factorizations of gene expression data guided by binding data.

Existing clustering methods do not deal well with overlapping clusters, are unstable and do not take into account the robustness of biological systems, or more complex background knowledge such as regulator binding data. Here we describe a nonnegative sparse factorization algorithm dealing with the above problems: cluster overlaps are allowed by design, the nonnegativity constraints implicitly approximate the robustness of biological systems and regulator binding data is used to guide the factorization. Preliminary results show the feasibility of our approach.

Functional discrimination of gene expression patterns in terms of the gene ontology.

The ever-growing amount of experimental data in molecular biology and genetics requires its automated analysis, by employing sophisticated knowledge discovery tools. We use an Inductive Logic Programming (ILP) learner to induce functional discrimination rules between genes studied using microarrays and found to be differentially expressed in three recently discovered subtypes of adenocarcinoma of the lung. The discrimination rules involve functional annotations from the Proteome HumanPSD database in terms of the Gene Ontology, whose hierarchical structure is essential for this task. While most of the lower levels of gene expression data (pre)processing have been automated, our work can be seen as a step toward automating the higher level functional analysis of the data. We view our application not just as a prototypical example of applying more sophisticated machine learning techniques to the functional analysis of genes, but also as an incentive for developing increasingly more sophisticated functional annotations and ontologies, that can be automatically processed by such learning algorithms.