Cytoscape: Getting Started Part 1

Cytoscape: It's installed, now what?
Intended audience:
Those who have already installed Cytoscape, have found the official Cytoscape tutorial, but still have little idea of what to do.

Background:
       I have noticed for some time, both my graduate colleagues and others that I have met at conferences seem to be quite annoyed with Cytoscape. A brief description of Cytoscape; it is a software commonly used to visually construct networks with nodes and edges (I am more interested in the biological networks). With the plethera of data out there, visualization is absolutely critical to helping other understand networks. In addition, Cytoscape can be used in conjunction with several "plugins" (which are basically functional Cytoscape software extensions) in order to calculate or better visualize any given network.

       As a sidenote, I will be using Cytoscape 2.8.3 for my tutorial. I suspect that, at least for all the core basics, different versions of the software will not vary too much.

Getting started 101:
Initially, this is (more or less) the software screen you get when you first open the application.

The first thing you want to do is import the data that will be used to construct your visual network. The great thing about Cytoscape is that it was built with Notepad or Excel in-mind. I am comfortable with Excel so I will use it to build my network consisting of protein-protein interactions (ie. I would have a list of which protein interacts with which protein).

Data import format:

The format is as such:

In this example, I have paired up my protein pairs such that I know SPAC1002.06C interacts with SPAC6G9.13C (the very first pair). Likewise, the second, third, and so forth lines represents other pairs. Cytoscape has a built-in feature which can easily remove all instances of duplicate pairs so if this is a problem, do not worry -- we will get to that soon!

Importing the data:

To upload your Excel file, go to FILE > IMPORT > Network from Table (Text, MS Excel)....

Make sure to select the appropriate columns for "Source Interaction" and "Target Interaction". Cytoscape also has an "Interaction Type" function in-case you have a file with multiple types of interactions (ie. a combination of protein interactions and genetic interactions). In our case, all pairs represent the same type of interaction (protein-protein). However, if multiple types of interactions are present and indicated by "Interaction Type", then different types of edges can be differentiated.

Importing attributes:

In Cytoscape, attributes are various descriptions of the nodes in your network. Let's say

Playing with the network:

This is the part that takes some time to explore.

Let Cytoscape organize your network-look in several ways:

Cytoscape allows you to manually drag nodes (the circles) and/or edges (the lines) however you want on the main screen. However, it is often best to let Cytoscape organize the "shape" of your network based on various built-in algorithms (more can be imported via plugins).

Starting with the upper toolbar, go to LAYOUT > Cytoscape Layouts. Choose any layout and see how it changes your network look. The layout seen above is Force-directed layout > (unweighted).

Zoom-in/ Zoom-out:

You can zoom in or out of the network by either scrolling with you mouse or using the zoom functions in the upper-left corner (the  two Magnifying Glasses).

In the lower left corner, there is a global view of your network with a blue rectangle over it. The blue rectangle tells you how much of your network you are currently seeing in the main  screen (the bigger area spanning the middle to upper-right of the screen).

The Right Lab, the Right Student, the Right Fit?

About a month ago, I had a casual conversation with a colleague that went something like this:

     "Colleague: You're an incoming 5th-year graduate student; you're just about done!"
      "Me: ...Actually, I'm an incoming 4th-year graduate student...."
      "Colleague: Oh! ...well, you've got a long way to go man!"

It was a friendly joke, of course, but it got me thinking: how do graduate students measure their time on this degree (PhD) and why? I have heard it measured in terms of years left. I have heard it measured in terms of years since started. I have even heard it measured in terms of "forever-ness". I have never really put too much thought into how long I have been here and how much more I have to go -- they have been merely numerical facts and averages. But thinking about it, many people seem to attach vast emotions to these numbers. But why?

In agreement with the famous PhD comics series, it seems that nearly all newcomers to the graduate program are "full of life, hope, and dreams of fame and success." It seems to usually be accompanied by the dream of "get rich quick." Finish the program asap and publish as many papers as possible. But I never really adhered to that dream coming into the PhD program; I entered hoping for a long tenure whereby I would emerge as a "Renaissance biologist." To me, graduate school is the last frontier of studentship -- where I would be most free to study all that I wanted (excluding my retirement where I could do just that too).

High-throughput versus low-throughput

I recently had a discussion with my lab-mate about the graduate student in a high-throughput versus a low-throughput lab. High-throughput labs run large-scale (often thousands to millions of mini-experiments), systematic experiments that usually feel repetitive. Publishable data almost always take a long time, even with maximum physical exertion, due to the scale of projects. However, once compiled, the end-data is always rich with information for a very wide-range of scientists. Low-throughput labs tend to be more focused onto a few genes/proteins; they are more detailed and thorough. Intellect, hard-work, and some luck can definitely lead to a slew of publishable work although this does not have to be the case.

By the end of this discussion, I concluded that both lab types are simply just that, different in nature; thus rejecting the notion of a better or worse work type.

This discussion ties into my earlier conversation because I came to a general hypothesis:
     "New graduate students tend to be overly optimistic about their future success as a graduate student
      and will tend to see low-throughput labs as the best opportunity to get rich quick [in papers]."

Biology is a fast-changing field and deciding how to make moves in the field will always be difficult. I have naively separated lab types into high versus low-throughput; clearly, there is a spectrum.For me, I will try to tread the unknown forest [high-throughput focus] and leave a path of my own. Only time will tell how that turns out.