First steps taken with our data set: let the searching for important genes begin!

Lab Note #1
May 21, 2014
It has been a while again since my last update, but generating the huge data sets you all so generously chipped in for takes some time and is frankly not that exciting to give you separate updates about. So, I decided to wait for the moment where I would actually start to work with them, and that moment has arrived! Therefore, expect some more frequent lab notes from now on. Especially because there are quite some other interesting things happening in our lab at the moment as well.  

As you might recall from my former lab note, the pilot experiments I ran looked really good, so I decided to go ahead and process all my actual samples. Just to remind you what they were again:
- Infected ant’s heads sampled while they were biting a twig, but still alive
- Infected ant’s heads that had died after they had been biting a twig
- A control for the fungus growing inside the infected ants, in the form of fungal growth outside its host
- A control for the sick ants, in the form of healthy ant’s heads
From all these samples I extracted RNA, which tells us which genes in the genome are used in different conditions. Now, the infected ant heads will of course contain RNA from both the fungus and the ant, and the method that was used to check the quality of the RNA (called a bioanalyzer) shows a different profile for insect samples as for fungal samples. However, the infected ant heads looked very much like the profile we get for the fungus by itself, giving the idea that there are mostly fungal cells!  

As a next step these RNA samples were used to make a library of all the genetic information they contain and after that the samples were run on a sequencer. What this sequencer does is reading the genetic information that is in the library in small fragments. These pieces of information are conveniently called reads. If there is more RNA present for a certain piece of information because it codes for a gene that is important, for instance in this case in behavioral manipulation, there will be more reads for this. The other way around, genes that are being used less will obviously have less reads. So counting all these reads will give an idea of what genes are up or down regulated during certain biological processes. Once all the reads are gathered, they need to be pieced together again, into what we call transcripts, to figure out what the genes are they are coding for, since the genes are much longer than the reads. This is computationally expensive so next to the sequencing taking some time, this needed some time as well.  

So, finally now the waiting is over and we can start to look at the enormous amount of data that is generated. I am getting help from some really skilled bioinformaticians for this and from one of the very talented grad students in our lab, Raquel Loreto. Since we are now getting information from both the ant and the fungal parasite we are splitting up the work. Raquel is going to look at the genes that are expressed in the ant and I am looking at the fungus. In order to make a start we first made a list of all the genes that are significantly different in the different samples compared to the other samples.  
Here are some rough numbers for you:  
For the fungus we find almost 7000 transcripts to be differentially expressed. From these almost 2000 are up regulated when the fungus grows inside the ant, and 5000 are down regulated.
For the ants we find that almost 4000 transcripts are differentially expressed in sick ant brains compared to healthy ones. Of these almost 1700 are up regulated and 2300 are down regulated. In fact we overall find way less transcripts for the ant genes in the sick ants, which again points out that they are almost mainly fungus by the time they are completely manipulated. How is that for a zombie analogy! These manipulated ants are almost completely fungus by the time they climb up to bite. Pretty amazing!  

Well, as you can see we are dealing with quite some large numbers of potentially interesting transcripts here, from which we now are trying to figure out what they code for. We already did a quick analysis to see what we might find and this pointed into some very exciting directions. But more about that in a future lab note because we first have to explore this further.  

On other news: one of our papers will be published soon! It’s an opinion paper discussing how experiments that look into gene expression involved in behavioral manipulation should be set up. Unfortunately, it won’t be published open access, but anyone who is interested in reading it, just let me know and I’ll send you a link or a pdf once it’s officially published!
Please wait...