Journal Review

Title	PLENTY OF FISH IN THE SEA: CORPORATE RESPONSIBILITY FOR OCEAN STEWARDSHIP
Journal Name	Harvard International Review
Volume and page	Volume 39.2 & 10 pages
Year	2018
Author	Darian Mcbain
Reviewer	Mikhael Hamashiah
Date	April 14 20 19
Research purposes	Save the ocean and the future of fisheries through corporate social responsibility
Research Subject	The ocean , fisheries and companies
Research methods	Case method
Operational Definition of Dependent Variables	Ocean stewardship
Ways & Dependent Variable Measurement Tools	The method used is to hold a meeting between sea food companies to show the current state of the marine ecosystem.
Independent Operational Definition	Corporate social responsibility
Research Steps	A meeting between companies is held to show the current ecosystem. After that it will be known that the company will contribute.
Research result	SeaBOS has brought together seafood companies to give them the responsibility of caring for and protecting fish and the sea
Strength of Research	-           This journal is able to explain how seafood companies can contribute to maintaining and preserving the sea through SeaBOS activities -           This journal has interesting images so that it can help in reading this journal so as not to get bored -           This journal is able to explain how the current sea conditions are
Research Weaknesses	-           This journal only explains that the company is able to help maintain the marine environment through CSR but does not explain whether the business is effective and efficient in maintaining the marine environment.
Conclusion	The ocean is a valuable resource and needs to be managed wisely. Sustainable fisheries management is an integral part of ocean management. Fishing is a practice that is almost as old as humans themselves, and on a commercial scale begins when there are many fish in the sea. But time changes. Fish stocks throughout the world are threatened.Anthropocentric pressure, such as climate change and pollutants, has been known for years. New and new issues, such as marine plastics and even more troublesome microplastic problems, are beginning to have an impact on marine health. To save the oceans, the responsibility of ocean companies requires the following: commitment to act for the common good, clarity of desired results, and collaboration with science, with regulators, and with civil society. Sustainability of the sea is no longer a problem of environmental campaigns. Scientists have published research on the decline of marine health for many years. The Economist will hold its fifth Fifth World Summit in 2018, bringing a discussion about how capital and the private sector can invest in marine sustainability. Instead, the United Nations held its first Sea Summit in New York in 2017. The private sector has recognized the risks and opportunities that may be far ahead of the government. SeaBOS is now running well in directly influencing the strategic direction and operational activities of participating companies, and is an exciting new opportunity for change. With businesses working together with science, governance and regulation, we may have hope for a sustainable marine future.

Title	Understanding Molecular Mechanisms of the Brain Through Transcriptomics
Journal Name
Volume and page	Volume 1 and 13 pages
Year	2019
Author	Wei Wang and Guang-Zhong Wang
Reviewer	Mikhael Hamashiah
Date	April 15, 2019
Research purposes	To research the brain using the Brain Transcriptome Analysis method
Research Subject	Mouse brain, brain of non-human primates and human brain
Research methods	Experimental method
Operational Definition of Dependent Variables	Methods of brain transcriptome analysis
Ways & Dependent Variable Measurement Tools	Analysis of Differential Gene Expressions The purpose of differential gene expression analysis is to detect changes in expression levels under different conditions using statistical methods. For microarray data, there are already established methods such as lymph (Ritchie et al., 2015), which use linear models to detect transcriptomic DE data, as well as to improve batch effects. For RNA-seq data, two models based on Poisson distribution and Negative Binomial distribution are often used (Sonesonand Delorenzi, 2013). Detailed comparisons of various methods, including well-designed R packages "DESeq" (Anders and Huber, 2010) and "edgeR" (Robinson et al., 2010), are discussed in previous reviews (Soneson and Delorenzi, 2013). The rapid development of high throughput techniques, such as microarray and NGS, makes it possible to assess the cell transcriptomic status at any given time (Barabási and Oltvai, 2004). Several methods are applied to analyze transcriptome data. Traditional methods involve comparison of knockouts with wild type samples, or diseases with a control group.Several pilot studies have provided the first glimpse of brain transcriptome, especially with the DE gene method, to compare KO with wildtype rats (Geschwind and Konopka, 2009).In the first step, DE analysis is performed, and the DE gene is identified. Furthermore, the functional annotation of the DE gene can be assessed by enriching gene ontology (GO) or KEGG pathway analysis, and enrichment of candidate gene genes can also be done.However, because the brain is a complex network system consisting of several types of cells, it is stated that analysis of ED may not be sufficient to obtain the structure underlying the gene expression data from the central nervous system (Miller et al., 2008; Oldham et al., 2008; Winden et al., 2009; Konopka, 2011). Network Analysis Network-based methods have proven to be more powerful than absolute quantities of expression levels, in expressing patterns of gene expression (Oldham et al., 2006; Miller et al., 2014; Hawrylycz et al., 2015, 2012), and have been found to be useful in analyzing work methods cell (Wang and Huang, 2014). By using network analysis, we can study the nature of high-level brain transcriptions. Data on gene expression profiles can be modeled as networks, where each gene corresponds to a node and gene pairs are connected by one side if the value of their expression is highly correlated (Parikshak et al., 2015). In networks, degrees are the basic characteristics of a node, and the degree distribution shows the probability that the selected node has the right N link. Degrees of nodes in random networks follow a Poisson distribution, while most biological networks approach free scale topologies, which means that fewer nodes are very connected and most nodes have low connectivity. Biological networks show high grouping features and consist of a set of modules, in which several nodes form densely connected communities that have less frequent connections with the rest of the network. In functional modules, cellular functions are executed by clustered molecules (Barabási and Oltvai, 2004). Co-expression of genes is defined as genes with the same expression pattern. Common measures of gene coexpression include Pearson correlation, Spearman correlation, Euclidean distance, and the angle between a pair of vectors observed (D'haeseleer et al., 2000; Horvath and Dong, 2008). In gene coexpression networks, modules refer to highly coordinated sets of genes (Barabási and Oltvai, 2004). To identify gene modules, several grouping methods have been used, including hierarchical grouping, model-based grouping, k-means, etc.(Figure 1). Genes in modules work together to achieve different functions.
Independent Operational Definition	Mechanism of brain molecules
Research Steps	Method of Brain Transcriptome Analysis Most methods of brain transcriptome analysis involve identification of genes that are expressed differently, either between normal tissue from various regions of the brain, or between normal tissue and tissue diseases such as autism or schizophrenia. The next step is to study the functions of genes expressed differently, as well as the properties of the network, such as its features in the expression network. The characterization of these traits lays the basis for understanding the role of these molecules in the brain. Analysis of Differential Gene Expressions The purpose of differential gene expression analysis is to detect changes in expression levels under different conditions using statistical methods. For microarray data, there are already established methods such as lymph (Ritchie et al., 2015), which use linear models to detect transcriptomic DE data, as well as to improve batch effects. For RNA-seq data, two models based on Poisson distribution and Negative Binomial distribution are often used (Sonesonand Delorenzi, 2013). Detailed comparisons of various methods, including well-designed R packages "DESeq" (Anders and Huber, 2010) and "edgeR" (Robinson et al., 2010), are discussed in previous reviews (Soneson and Delorenzi, 2013). The rapid development of high throughput techniques, such as microarray and NGS, makes it possible to assess the cell transcriptomicstatus at any given time (Barabási and Oltvai, 2004). Several methods are applied to analyze transcriptome data. Traditional methods involve comparison of knockouts with wild type samples, or diseases with a control group.Several pilot studies have provided the first glimpse of brain transcriptome, especially with the DE gene method, to compare KO with wildtype rats (Geschwind and Konopka, 2009).In the first step, DE analysis is performed, and the DE gene is identified. Furthermore, the functional annotation of the DE gene can be assessed by enriching gene ontology (GO) or KEGG pathway analysis, and enrichment of candidate gene genes can also be done.However, because the brain is a complex network system consisting of several types of cells, it is stated that analysis of ED may not be sufficient to obtain the structure underlying the gene expression data from the central nervous system (Miller et al., 2008; Oldham et al., 2008; Winden et al., 2009; Konopka, 2011). Network Analysis Network-based methods have proven to be more powerful than absolute quantities of expression levels, in expressing patterns of gene expression (Oldham et al., 2006; Miller et al., 2014; Hawrylycz et al., 2015, 2012), and have been found to be useful in analyzing work methods cell (Wang and Huang, 2014). By using network analysis, we can study the nature of high-level brain transcriptions. Data on gene expression profiles can be modeled as networks, where each gene corresponds to a node and gene pairs are connected by one side if the value of their expression is highly correlated (Parikshak et al., 2015). In networks, degrees are the basic characteristics of a node, and the degree distribution shows the probability that the selected node has the right N link. Degrees of nodes in random networks follow a Poisson distribution, while most biological networks approach free scale topologies, which means that fewer nodes are very connected and most nodes have low connectivity. Biological networks show high grouping features and consist of a set of modules, in which several nodes form densely connected communities that have less frequent connections with the rest of the network. In functional modules, cellular functions are executed by clustered molecules (Barabási and Oltvai, 2004). Co-expression of genes is defined as genes with the same expression pattern. Common measures of gene coexpression include Pearson correlation, Spearman correlation, Euclidean distance, and the angle between a pair of vectors observed (D'haeseleer et al., 2000; Horvath and Dong, 2008). In gene coexpression networks, modules refer to highly coordinated sets of genes (Barabási and Oltvai, 2004). To identify gene modules, several grouping methods have been used, including hierarchical grouping, model-based grouping, k-means, etc.(Figure 1). The genes in the module work together to achieve different functions . One of the main objectives of the analysis of joint expression networks is to identify gene modules (Barabási and Oltvai, 2004). The pattern of co-expression of brain genes is mainly evaluated by correlation-based measurements (Mahfouz et al., 2017). By detecting gene expression patterns similar to disease genes, silico predictions can be made using the gene expression approach. To find groups of genes that are expressed together in a set of samples, the unsupervised method commonly used is grouping hierarchies (Mahfouz et al., 2017).One method used to identify shared expression modules is Pearson correlation, the most popular measure of expression expression(Wang and Huang, 2014), as distance measurement for hierarchical grouping. Hard thresholding is then applied to produce tissue(Li et al., 2016). One method that is widely used for network co-expression construction is weighted correlation network analysis (WGCNA), which was first introduced by Zhang and Horvath (2005 ). This is an informative method for detecting biologically relevant patterns using high dimensional data sets, and allows for assessment of the relationship of modules with experimental properties (Zhang and Horvath, 2005). Genes with strong covarying patterns are grouped into modules in the sample set. The identified modules are marked with eigengena modules, and the hub genes refer to genes that are highly eigengene correlated. WGCNA is a biological system method used to build gene expression modules along with an unsupervised grouping approach and has been widely applied for the analysis of mammalian brain transcriptomes (Oldham et al., 2006; Hawrylycz et al., 2012; Thompson et al., 2014 ; Bakken et al., 2016). WGCNA looks for gene expression modules along with high topological overlap(Zhang and Horvath, 2005). First, soft threshold strength is chosen to calculate proximity, which is then transformed into topological overlapping matrix. Then, gene dendrogram can be produced through hierarchical grouping. Finally, modules are identified using the dynamic tree cutting method to cut off branches (Langfelderand Horvath, 2008). In addition to the popular WGCNA, there are also a number of different methods that have been developed for cluster analysis and subsequently detect analysis of network modularity (Figure 1). The K-mean clustering method specifies the number of clusters (K) before clustering, and then, based on calculation of distances (usually Euclidean distances), all different modules are detected(Jain , 2010). However, different cluster initialization can cause different final groupings.Another reasonable approach is based on the probability model, network nodes that are calculated based on the probability distribution of genes (Yeung et al., 2001). Model-based methods can capture correlations and dependencies between attributes, and are implemented in the "clcl" R package (Yeung et al., 2001).
Research result	-           Mouse brain Using data on voxel expression, Thompson et al. (2014) exploredtemporal coexpression patterns of rat brains in diencephalon for three time periods: "embryonic," "postnatal" and "all." They analyzed the "all" period and found that the genes in the two modules showed strong regulation at diencephalon at P14 and P28. They further examined the postnatal cluster and found that a famous set of oligodendrocyte genes was not widely distributed until P14. The most interesting pattern of temporal expression is that P14 shows a strong thalamus specific expression of the dominant TF gene. The authors concluded that this might coincide with the opening of the eye and the initial reception of visual stimulation by the thalamus. -           The brain of a non-human primate Explore patterns of spatio and temporal expression of the postnatal brain of rhesus monkeys. Five brain regions are considered for gene-wide expression of the genome at birth, infants, childhood and young adults. They identified 27 modules in total. Correlating with each module eigengene with age and brain region, they found several modules related to age, with a gradual shift in postnatal gene expression. They also identified specific expression modules for cortical areas such as the primary visual cortex enrichment module (M6).They explored the expression of the M6 ​​gene, and confirmed previous findings that, in rhesus monkeys and adult human brains, patterns of gene expression in the primary visual cortex differ from other brain regions. However, in the cerebral cortex, there are striking differences between humans and chimpanzees, consistent with cortical expansion in the human lineage (Oldham et al., 2006). In addition, Zhu et al. (2018) compared the development of the nervous system between humans and apes, and detected a pattern of cup-shaped transcriptriptic differences between the two species. In addition, they also identified different-human gene expression modules, which showed differences in molecular mechanisms for species divergence, which could play a role in mental disorders. Therefore, to uncover the special features of the human brain at the molecular level, it is necessary to use transcriptomes of the human brain rather than transcriptomes of the brains of nonhuman primates. -           Human brain The development of the human brain is a complex process and depends on the regulation of proper gene expression (Rakic, 2009; Rubenstein, 2011). Using transcriptome data from highly differential stability genes, Hawrylycz et al. (2015) built a consensus gene expression network and found several modules with annotations related to the most neuronal functions. Allocating genes to each module identified according to the gene correlation with the appropriate eigengene modules, they detected a number of highly selective modules for certain brain regions. Interestingly, when assessing the preservation of modules between humans and mice, they found that some modules related to neurons were well maintained, whereas many of the most non-neuronal modules were poorly maintained. However, several genes differ in their expression that are patterned across species even in highly maintained modules. Modules related to neurons are better stored than modules associated with glia. Using data from 16 regions consisting of six brain structures during the period before and after birth, Kang et al. (2011) created a gene expression network and identified 29 modules related to different spatio-temporal profiles. They found that 90% of the expressed genes are regulated differently on all transcripts or exon levels in all brain regions or periods of brain development. Among these modules, the M8 shows the highest expression level in the initial fetal neocortex and hippocampus, and then a progressive decrease in expression levels to the baby. The gene hub of M8 is involved in the development of neocortex and neuron projections of the hippocampus.In addition, they identified two temporarily arranged modules, with opposite developmental trajectories: M20 showed a decrease in expression while M2 showed increased expression, with a shift just before birth, which suggests that environmental influences might be related to transcription changes during this period of brain development . . Furthermore, using gene expression data from 11 areas of the neocortex in the human and ape brains, Pletikos et al. (2014) analyzed patterns of spatial expression between regions throughout the development period. They first applied the ANOVA approach to identify genes expressed differently between neocortical regions, in each development period and propose an hourglass model of interareal transcriptional divergence over time, suggesting that the spatial patterns of interareal divergence were mainly driven by a number of functional areas.In addition, to gain insight into the organization of neocortical transcriptomes, they then carried out WGCNA with samples from two periods (periods of fetal development and from adolescence onwards) increasing interarral differences and each identifying 122 modules and 207 modules. Most fetal modules show a pattern of temporarily determined area and lose the difference in prominent areas after birth. In contrast, adolescent and adult modules are more stable over time, and show less complex spatial patterns. In addition, Li et al. (2018) integrated transcriptome, DNA methylation, and histone modification data from 16 brain regions, and revealed cup-shaped patterns of regional differences during prenatal and postnatal development. In particular, they identified a group of gene expression modules related to dynamic spatiotemporal trajectories and found that many modules were enriched with specific cell types or disease-related genes. Using organoids from human pluripotent cells, Amiri et al. (2018) modeled cerebral cortical development between 5 and 16 weeks after conception. They identified gene networks and enhancement modules and found that several enhancement modules met with gene modules, showing the regulation of genes that are expressed together by increasing cross-time. -           Mental disability Integrating analysis of expression networks into traditional differential gene expression analysis finds features of normal mammalian brains and broadens our knowledge of spatio-temporal events in the development of brain mammals over the past decade. In addition, to reveal the molecular mechanisms of neuronal disorders, such as Autism spectrum disorder (ASD), Alzheimer's disease (AD), Schizophrenia, etc., joint expression tissue is applied to compare healthy and diseased brains, which will also reveal important biological pathways in disorders this and provide potential biomarkers or therapeutic targets (Keo et al., 2017; Seyfried et al., 2017; Mostafavi et al., 2018; Rajarajan et al., 2018). -           Utilizing gene expression analysis to decode the mental disorder mechanism is a powerful tool because it is large-scale, high throughput and cost-effective. ASD is a group of neurodevelopmental disorders characterized by deficits in social functions and repetitive and limited behaviors or interests (Bourgeron, 2015). Previous findings indicate that the ASD gene is only enriched in the pathway during early fetal development (Parikshak et al., 2013). In postnatal rhesus brain tissue, Bakken et al. (2015) found that the ASD-enriched module showed significant enrichment in the neocortex. Gene expression in one of these modules is high in the neonatal cortex and striatum but low during the period of development of infants and adolescents. Combining solid temporal samples of the prenatal and postnatal periods, Bakken et al. (2016) showed a high-resolution transcriptional atlas of ape brain development (Macaca mulatta) with good anatomical division of cortical and subcortical regions associated with human neuron disease.They found that many ASD genes showed coordinated expression in postmitotic neurons both before and after birth. They also found that in a neuron-enriched progenitor module, the MCPH gene was enriched at the early to mid-age of the fetus. There is no gene enrichment related to intellectual disability observed in any module. Using 109 cortical miRNA samples, Wu et al. (2016) applied WGCNA and identified 11 modules. By examining the relationship between eigengene modules and the properties of ASD, they detected three modules that significantly correlated with ASD, and successfully predicted and validated two transcription factors that regulate neuronal genes in ASD. -           Alzheimer's disease is the most common cause of neurodegenerative dementia (Verheijen and Sleegers, 2018). Using 19 cortical regions, Wang et al. (2016) build a network of regional-specific shared expressions, and shared expression modules ordered by rank and brain region based on their relationship with AD pathological properties. They found that the temporal lobe gyrus showed the largest and earliest gene expression abnormalities. Mostafavi et al. (2018) apply network-based methods and identify specific genes that are associated with properties related to AD. By integrating clinical, neuropathological data and gene expression, they detected a expression module associated with cognitive decline and β-amyloid load. In addition, they identified two genes in the module, INPPL1 and PLXNB1, as potential AD therapy targets. -           Gandal et al. (2018a) analyzed the transcriptome of five major psychiatric disorders, including ASD and schizophrenia, and identified a number of shared coexpression modules and specific disorders. They found the regulated modules, which are related to astrocytes, and some downregulated modules, which are annotated as neuronal or mitochondrial, throughout ASD, schizophrenia, and bipolar disorder, suggesting the molecular convergence pathway of major neuropsychiatric diseases. -           Nevertheless, the PsychENCODE consortium integrates multiomic dataand provides comprehensive resources for functional genomics of the human brain (Wang et al., 2018). For example, Gandal et al. (2018b) integrates RNA-seq and genotypes in brain samples with ASD, schizophrenia, and bipolar disorder, and detects gene expression modules associated with each disorder.They found that one module, which was associated with microglial cell markers, was set up in ASD, and was regulated downward in schizophrenia and bipolar disorder, suggesting previously uncovered nerve mechanisms. -           Integration of expression data with clinical traits allows identification of new disease-related modules and hub genes, which provide potential therapeutic targets for related neuronal disorders.
Strength of Research	-           This journal is able to explain how research uses the transcriptomic method in mice, non-human and human primates so that readers can understand how brain functions can work and it is understandable why every brain in living things has a different function. -           The journal also explains how this method applies to the brain that has a mental disability and finds many interesting and new facts
Research Weaknesses	-           The weakness of this journal is thatthis journal can not provide insight into the behavior of various cell types, which is an important aspect of brain research.Likewise, the analysis methodology developed for mass samples may not be suitable for analyzing single cell data with network algorithms. In this mini review, newly emerging single cell sequencing data are not included due to space constraints.
Conclusion	Transcriptomic data from the mammalian brain provides a great opportunity to illuminate how the brain works at the molecular level. The current status of this field has given us great insight into the patterns of development of brain molecules, and we expect more primate brains to be included in future research. In addition, other molecular activities such as microRNA and non-coding RNA must be profiled at the brain scale as well. In this article, we summarize the progress made by various researchers in brain transcriptome analysis in recent years. In addition to traditional DE analysis, network-based methods offer an unattended perspective for analyzing large-scale data from the mouse to the human brain, as well as data on the various stages of development of each species. In addition, system level analysis assembles a single gene correlation and enables the discovery of the main pathway. As the rapid development of NGS in the past decade has accelerated brain transcriptomic research, the knowledge gained from this field can facilitate the decomposition of brain complexity and help us gain valuable insights about the regulation of brain function.However, the use of network-based methods that are integrated with clinical features and experimental validation Transcriptomic data from the mammalian brain provides a great opportunity to illuminate how the brain works at the molecular level.   The current status of this field has given us great insight into the patterns of development of brain molecules, and we expect more primate brains to be included in future research.   In addition, other molecular activities such as microRNA and non-coding RNA must be profiled at the brain scale as well.   In this article, we summarize the progress made by various researchers in brain transcriptome analysis in recent years.   In addition to traditional DE analysis, network-based methods offer an unattended perspective for analyzing large-scale data from the mouse to the human brain, as well as data on the various stages of development of each species.   In addition, system level analysis assembles a single gene correlation and enables the discovery of the main pathway.   As the rapid development of NGS in the past decade has accelerated brain transcriptomic research, the knowledge gained from this field can facilitate the decomposition of brain complexity and help us gain valuable insights about the regulation of brain function.  However, the use of network-based methods is integrated with clinical features and   validation  experimental. One of the main limitations of the analysis of mass transcriptomome samples is that it cannot provide insight into the behavior of various cell types, which is an important aspect of brain research.   Likewise, the analysis methodology developed for mass samples may not be suitable for analyzing single cell data with network algorithms.   In this mini review, newly emerging single cell sequencing data are not included due to space constraints.   Analysis of genes expressed differently between different cell types or markers between different cell types will be an important topic in the future.  Research in this field is progressing rapidly , and we are looking forward to some important improvements to identify cell types associated with genes that are expressed differently in the future.

Title	CORPORATE AND SOCIAL RESPONSIBILITY
Journal Name	Journal of Property Management
Volume and page	Volume 83 .6 and 1 page
Year	2018
Author	Association of Realtors
Reviewer	Mikhael Hamashiah
Date	April 15, 2019
Research purposes	Give examples of corporate social responsibility to the community
Research Subject	Cristo Rey Tampa High School
Research methods	Historical method
Operational Definition of Dependent Variables	Social responsibility
Ways & Dependent Variable Measurement Tools	Visiting the place of company responsibility
Independent Operational Definition	Corporate
Research Steps	They visited the CSR place and then they researched the CSR location.
Research result	R esp rop carries out corporate social responsibility by rebuilding the high school Cristo Rey Tampa for underprivileged students so that they are able to obtain education for the world of work later
Strength of Research	With this journal, readers can find out what Resprop has done to implement their social responsibility
Research Weaknesses	Journals do not tell whether what they have done has the effect that the company expected.
Conclusion	Responder companies can provide assistance to the community for ongoing support from schools including fundraising, managing renovations, teaching, guidance and job training. It is hoped that the lower class will grow and develop. And resprop believes in being able to overcome problems by abandoning thoughts that inhibit them to develop and show that everyone can do great things.