>> Hello and welcome to the first lecture of the course, Integrated Analysis in Systems Biology. I'm Susana Neves, Assistant Professor at the Icahn School of Medicine at Mount Sinai and I'll be your instructor for this course. The goal of this course is to prepare you for the capstone project and to help you apply the computational approaches that you have learned in our other courses, such as Network Analysis and Systems Biology by Abby Moyan and Dynamical Modeling Methods for Systems Biology by Eric Sobe. This course will consist of director readings of three original research articles in systems biology. For each article, we will cover the biological background, the experimental and computational approaches used and key results. At the end of each article, there will be a problem set and structured discussions. We're going to start today with an overview of the general features of a systems biology project. We will also cover all the sections of a research article. The Scope of Systems Biology is the comprehensive understanding of the functioning of cells, tissues and ultimately organisms. Systems biology can be divided into two broad approaches to study this biological complexity and these approaches are complementary, but are different in scope and scale. In one end, we have the Omics Tradition that consists of genomic data mRNA expression data and protein-protein interaction data. And all of this type of data can be analyzed using statistical base or graph theory base approaches. At the other end is the Physiology or Biochemistry Tradition that provide us with quantitative data sort of, such as signaling and electrophysiology data that can be used to develop dynamical models. Each approach has its own benefits and limitations. With one providing us with large scale, high-throughput unbiased data, but at the expense of being descriptive and qualitative. On the other hand, dynamical models can be very mechanistic and quantitative, but they tend to be small in scale and very low-throughput. So how do we design a systems biology study? First, one must survey the literature and determine what is known. And more importantly, identify what are the gaps in our current understanding that need to be addressed? Once we have this information, we can formulate the objective of the study or the research question. One can go further and frame the biological question as the hypothesis. And one must ensure that the hypothesis is clear and concise, that it's experimentally testable. And more importantly, that it's consistent with all the knowledge that we have available. >> For instance, if one is interested in identifying novel therapeutic targets to treat cancer. The objective of the study could be to identify what genes are regulated in a cancer tumor. And perhaps, that could help us identify upstream signaling that we could target. >> Next, we need to determine the scope of the study in order to ensure that the biological question can be accurately answered from the experiments and analysis plan, then we must decide on the scale of the data gathering, that either we perform comprehensive or global gene expression analysis or we can be targeting. Meaning that we just choose a subset of genes that may be regulated by a specific path. >> So we have to identify the appropriate biological system. For instance, can we address the research question using just associated cells or do we need tissue or whole animal data? Next, we have to decide on what are the relevant experimental computational and statistical methods in order to gather the data and analyze the data? So for example is it appropriate to use mRNAseq or a micro array data? Or are we just looking at subsets of genes that were Interested in. So maybe a quantitative PCR may be appropriate. And then we have to decide whether we're doing a one time point or we're doing multiple time series. And so where we're looking at the evolution of these changes or are we just doing a single comparison. >> Next, we need to formalize the details of the experimental computational design. >> So what are the appropriate biological and quality controls? So, if you're doing a comparison study, what is the appropriate comparison? What sort of perturbations the variations that you use, the different drug treatments, different mutations that you should compare? Finally, if, what should be your experimental outputs? What is exactly that, that your target should be? If you're doing a time course how often should you have a time point? What would be appropriate? And also what would be the sample size to satisfy a statistical requirements? And finally, there's several replication criteria and there's two types of replicates biological in technical replicates. There's biological replicates that account for the variability that exists in the biological system and those kind of replicates are basically, for example, doing multiple animals in the same treatment. While there's technical replicates that deal with variability that is inherent to the technology used to acquire the data. So that would be having a single sample that you basically measure multiple times in the same machine to see how variable the machine is. Once we collected the experimental data we have to determine whether the results are significant using the pre-determined statistical analysis and in certain cases further data processing and analysis is required. And these data analysis times can lead to predictions that can be further experimentally validated. And that's something that we will see in the next lecture. The last steps would be to assess whether the results are supportive of our initial hypotheses. If they're not, we must either reconsider our hypotheses or revise our scope of our study. Well, perhaps the terminate, alternative experimental computational approaches would be more appropriate. And finally, after all the effort required for the design, data collection and data analysis, publication is a crucial endpoint to share the results with the broader scientific community. >> Next, we're going to discuss the section in a research article. All research articles have the same standard format, consisting of several sections such as abstract introduction, methods, figures and tables, results and discussion. >> And just to illustrate these sections, we will use the first research article that we will discuss in DevNet. Titled, a systems approach identifies HIPK2, as a key regulator of kidney fibrosis. So the first section is the abstract. The abstract is basically the summary of the study and it should accurately reflect the content of the research article. And as you can see from the abstract, the first sentence is about brief introduction to the problem. This is followed by a somewhat detailed summary of the approach used and also of the results obtained and the abstract, it ends with a brief conclusion. The next section is the introduction. And the introduction is important, because he provides all the background information to justify the need of the study. So, in this case, the first paragraph includes all the current available knowledge that by a logical problem that they're interested in. The last sentence is particularly important as it identifies a gap in the current understanding that needs to be addressed. >> The next two paragraphs introduce the approach that was used and also the objective of the study. So the, the key points in the introduction should be what is known about the topic? Why is it important? What is exactly the, the, the problem that, that needs to be addressed? And what is the hypothesis? And what approaches are being used to address the the problem? >> Following the introduction, there are generally three major components in the research article. Methods, results and discussion. But the order of the sections and whether they are all separate will depend on a particular journal. Some journals format includes conclusions at the end of the discussion, others have a separate section. Some have methods followed by results, followed by discussion. And in some cases, results and discussions are combined into a single section. The research article we're discussing next time places the methods at the end of the paper. >> The next section is the methods section. In, in here, you are expected to describe all the methods, both computational and experimental used in the study with enough detail to be able to replicate all the findings. And so, as you can see in the text there is a significant amount of, for example, for the experimental reagents the companies where they were purchased or provided. Concentrations that are provided. Statistical analysis that are described. And there is even more descriptions of the computational analysis into supplemental section. The next figures and tables are a component of the paper and they're generally part of the result section and they always contain a legend that is descriptive and at times may include some statistical significant measures. The results section presents the data that you have collected in a rational way without excessive interpretation, just enough to make the presentation logical and connected. >> In this section each of the figures and tables should be referenced with a summary statement of what is demonstrated in each. >> A after all the results have been described, there is a discussion. The discussion is basically a brief summary of the key results. And the key here is to give the authors the opportunity to place their results within the current literature. Stressing the novelty of the findings acknowledging the limitations speculating about the findings, since summarizing possible potential implications and also future directions. >> The last paragraph of the discussion should be com, a conclusion paragraph or at least a couple of sentences. And in certain in certain journals the format includes a defined conclusion section. But in, in, in this one, the discussion and the conclusion are combined. So with this, we're going to end the first lecture and the course. The next lecture, we will cover the first research article of the course. Thank you.