Sample Sequencing

During the last two weeks, students have been submitting samples of their sourdough starters for DNA sequencing.  We will be getting data and results back soon, but for now, we will be practicing DNA analysis using data from last semester's sourdough starters. 

What are we sequencing? 

The purpose of this research project is to determine the exact sequences of nucleotides within the sourdough starters to figure out which microorganisms are present. We will be sequencing a few very specific genes or regions within a genome of  many different organisms. There are two genes that we will be sequencing for this set of sourdough starter samples; the 16S rRNA gene, which is present in bacteria and the ITS region, which is present in eukaryotic organisms such as fungi. 

Shotgun vs Amplicon-based Metagenomic Sequencing

We discussed two ways that we could sequence our sourdough starter samples, each having their own pros and cons. Shotgun metagenomic sequencing involves sequencing all of the genomic DNA from all organisms in the sample. Amplicon-based metagenomic sequencing involves the sequencing of specific regions such as the 16S gene and ITS region. I assumed that there would be differences in the costs of these different types of sequencing as well as how appropriate each one would be for different purposes. Depending on what you want to sequence, one type may be more affordable than the other and easier to use.

After doing a bit of research on the differences between these two types of sequencing, I found a really helpful resource that provides a table comparing and contrasting the two types. The description of the table states, "Overall, shotgun metagenomic sequencing has greater taxonomy resolution, functional profiling, and cross-domain coverage. In all other aspects, including price and sample origin compatibility, 16S/ITS sequencing has the advantage." I also found that Amplicon-based metagenomic sequencing may be more useful when trying to determine which bacteria and fungi are present because there is much better coverage of the 16S/ITS databases. 

Depending on what you are sequencing, you can learn more from one type than the other. For example, if you are sequencing a sample containing DNA from a human microbiome or functional profiling of your sample is required, it would likely be more beneficial to use shotgun metagenomic sequencing. This type of sequencing allows metabolic function analysis. Shotgun metagenomic sequencing can also be used to achieve a higher taxonomic resolution than can be achieved by Amplicon-based metagenomic sequencing.

Although it does seem like there may be more time and steps involved when using Amplicon-based metagenomic sequencing, I believe that this type is better for the purpose of our project. It is more cost-efficient and is capable of helping us determine the genus and species of different microbes. 

QIIME

This was my first time using QIIME to graph and analyze data. I found that it was relatively easy to use and had various features that made the analysis pretty interesting. I really liked being able to easily switch between different taxonomic levels and ways of sorting the data. For my two graphs on the 16S data and ITS region data, I changed the settings to level 7 taxonomy and sorted the data by treatment groups to compare the results from different fruits and control samples. 

While using QIIME, I did not encounter any difficulties. The types of files used to enter the data and create the graphs were hard to work with, but after converting them to files more compatible with my computer, it was easy to open and upload them. One thing I wished it could have done was show the sample number each time no matter how the data was being sorted. I think it would have been easier for me to make connections if I could have seen which sample was which each time. For example, I wish it would have named them like "apple, 28" or "control, 104." It would have been great to have also seen these same numbers while looking at the name of the individual that submitted the sample. Overall, I thought the process was pretty simple. 

This is a graph I created on QIIME for last semester's 16S data. I changed the settings many times in attempt to find more distinct patterns within the graph. The final graph that I am showing above is a depiction of level 7 taxonomy sorted by treatment group. As you can see, each type of fruit and then all control samples have been sorted into groups. This allowed me to compare samples that have similar contents. Although I could not find any distinct patterns, one thing I did notice was how much the similar types of starters could differ. For example, if you look at the legend below and then at the blueberry sample within the graph, you can see that one blueberry sample contained approximately 95% Lactobacillus (purple) while the other blueberry sample contained less than 10% of that same species of bacteria. I thought these types of findings were really interesting and could possibly raise a lot of questions.  

16S Data Legend

Below you will find a graph of the ITS data from last semester along with its legend. 

Here is the graph of the ITS data I created on QIIME. Similar to the first graph, this graph depicts level 7 taxonomy and is sorted by treatment groups. Referring to the legend below, you can see that a large portion of the microbes in this dataset have been labeled "unassigned." These microbes could not be identified in the database, which shows us just how extensive the list of known bacteria is compared to that of the fungi. Regardless of how I arranged or sorted the graph, I could not find any obvious patterns or trends. However, it was apparent that there was a larger amount of Saccharomyces in most of the samples than any other identified fungi. 

ITS Data Legend

Potential Research Questions

1) Question: Do the control starter samples and experimental starter samples grown by the same person have similar microbial compositions? 

Explanation: I believe answering this type of question may help determine if the environment in which the sourdough starters were grown may have influenced the microbes that were growing within the samples. It should not be too difficult to compare the microbial composition of the control sample from one person to the composition of their experimental fruit sample(s). 

2) Question: Are there any distinct differences in microbial compositions of experimental starters that contain fresh apples vs those containing applesauce, or are they similar? 

        

Explanation: Although my starters did not grow well enough to be submitted for sequencing, I would like to possibly ask and answer some questions that I had relating to my own starters if others in the group used the same fruits. My original experimental starter contained applesauce and my second experimental starter that I began later on contained an apple slice. That apple came from the grocery store. I was curious to see if there would be any differences between these two starter's microbial compositions since they are from the same type of fruit that has just been prepared differently. I was wondering if there were any differences between a fresh apple and applesauce that could lead to different phenotypic characteristics and types of microbes present. 

3) Question: Does the age of the sourdough starters have an impact on the diversity or abundance of microorganisms present within the samples? 

Explanation: This question could possibly be answered in a few different ways or on various scales. You could look for differences just in the bacteria present or just in the fungi present. Or you might be able to compare those two major groups. I might be able to ask "Do older starters have more identifiable fungi than younger starters?" Questions could also be more specific and ask about the relative frequency of one species in particular for starters of differing ages. I think it might would be a good idea to come up with more specific questions like this once I have seen the actual data we will be using. The more specific questions may help answer the bigger question that has been asked above. 

Citations

https://view.qiime2.org/

https://www.zymoresearch.com/blogs/blog/16s-sequencing-vs-shotgun-metagenomic-sequencing