Lab 8: Modified Alexander et al. tube protocol for DNA extraction

October 22, 2019

In this lab, we were given the sample of the Mimulus guttatus leaf that we had picked in the previous field trip that we had taken. We also were given two more samples of leaves that Alec had stored for us. When I was given the leaves, they were dried up in the tubes full of silica. Before we began using the leaf samples, we labeled 3 2.0 mL tubes with our samples codes. The first sample was labeled ‘RDRK 001’ the second was ‘EB’ and the third sample was labeled ‘SHOR 005.’  In each tube, I added three sterile 3.2-mm stainless steel beads to each tube. I added a small amount of leaf tissue to each tube using clean tweezers between each sample to avoid cross contamination.

I capped my tubes and loaded them within a tube rack into the modified reciprocating saw rack and mounted the rack to the saw. Professor Paul made sure that the blade was completely secured and locked into the saw before plugging in the saw. Professor Paul put safety goggles on and turned on the reciprocating saw on speed 3 for 40 seconds.

We briefly spun down the tubes in the centrifuge for 15-20 seconds at fast speed to pull plant dust down from the lids. After the plant dust was pulled down, I added 434 μL preheated grind buffer to each tube. We incubated the buffer grindate at 65°C for 10 minutes in a hot water bath. We made sure to mix the tubes by inverting them every 3 minutes. After incubation, we took the tubes out and added 130 μL of 3M pH 4.7 potassium acetate into each tube. The tubes were inverted several times and incubated on ice for 5 minutes. After ice incubation, we took the tubes out and placed them in a centrifuge at maximum force for 20 minutes making sure that the centrifuge was balanced.

We labeled new 1.5 mL tubes with the sample ID. I transferred the supernatant to these sterile 1.5 mL microcentrifuge tubes avoiding the transfer of any precipitate. I added 1.5 volumes of binding buffer.

We applied 650 μL of the mixture from the previous step to the Epoch spin column tubes and centrifuged the tubes for 10 minutes until all the liquid has passed through. The centrifuge was placed at 15,000 rpm and the flow-through was placed in a hazardous waste container once the centrifuge was complete. This step was repeated once more using the remaining volume from the previous step.

We washed the DNA bound to silica membrane by adding 500 μL of 70% EtOH to the column and centrifuging it at 15,000 rpm until all the liquid passed through to the collection tube. This process was done in 8 minutes and the flow-through was discarded. Repeat this step once more.

After discarding the flow-through from the last step, we centrifuged the columns at 15,000 rpm for an additional 5 minutes to make sure that any residual ethanol was removed. We were instructed to discard the collection tubes and place the columns in sterile 1.5 mL microcentrifuge tubes and add 100 μL of pure H2O. However, I made the error of placing the centrifuged columns into the tubes that contained the binding buffer and place 100 μL of H2O. When I caught my mistake, Professor Paul advised me to transfer the columns into clean 1.5 mL tubes and add an additional 100 μL of 65°C H2O to each tube and let it sit for 5 minutes. After this was done, I put the tubes in the centrifuge for 2 minutes at 15,000 rpm to elute the DNA. The new tubes were labeled ‘EB 1’ ‘EB 2’ and ‘EB 3’. I drew a star on the tubes that the columns were initially put into so that we may use them in the future if we find that no DNA was extracted into the new tubes.



Lab 7: Phylogenetic Inference

October 8, 2019

In Lab Portion:

To prepare for this lab, we were assigned the task of assembling an alignment of COI sequences that includes our fish DNA barcode sequences and sequences downloaded from NCBI through Geneious. Once we had this alignment, we were told to use the ‘allow editing’ tool to ‘clean-up’ our alignment so that all the sequences begin and end at the same point. When I looked through the first 20 columns of my alignment, I found only one column that showed a polymorphism.

Choosing the best model of molecular evolution

Next, we used a program called jmodelTest2 to establish the best model of molecular evolution for my sequences. Within the folder called “jmodeltest-2.1.10.tar.gz” there was a file named ‘jModelTest.jar’ that opened a window. Before using the jmodelTest2, we had to return back to the Geneious program and export the alignment in Phylip format (relaxed). Once that was completed, I could go back to the jModelTest2 and open the exported alignment. We then were able to compute the likelihood scores’ of each model of molecular evolution as the program goes through a set of hierarchal set of 88 models based on an initial tree that it uses for all tests.

Once the likelihood calculations were completed, we used two methods to choose the best model based on some optimality criterion. The two methods were the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). The best model based on AIC was model 79. Similarly, the best model based on BIC was model 79.

Bayesian Inference

Back in Geneious, we were able to install the ‘Mr.Bayes’ tab by using ‘Tools’ > ‘plugins’ in the menu bar. We selected our alignment and chose Mr.Bayes as our ‘tree’ of choice. In the settings, I chose to use the GTR (complicated) substitution model as my model of molecular evolution that was closest to my jModelTest complexity. Next, I chose to set my ‘rate variation’ to ‘gamma’ because my jModelTest best model contained a ‘+G’ in it. The outgroup that I chose for this alignment was a shark species under the consensus identity #HM991188. Under the MCMC settings, I chose a ‘chain length’ of 10,100, a ‘subsample frequency’ of 200, and a ‘burn-in length’ of 100,000. Under the ‘Priors’ settings, we used ‘Unconstrained Branch Length: Exponential (10)’ and ‘Shape Parameter (10)’ which affects the shape of the prior statistical distribution. The Posterior distribution showed that this analysis was not run for long enough. The Trace figures showed an upward sloping line that is representative of an increase in the posterior probability and is part of what you want to remove with the burn-in length. We ran our analysis one more time but used a ‘chain length’ of 110,000 and 10,000 for the ‘burn-in length’ this time. With this second run, we found that the Posterior distribution was normal and the Trace figures looked like the ‘fuzzy caterpillar’. I recovered the same clades in my tree and they both had similar support values.

Maximum Likelihood 

In this last section, I used maximum likelihood to infer a phylogenetic tree of my aligned data set. First, I went the menu bar and through ‘Tools’>’Plugin’, I was able to install RAxML which does a very fast maximum likelihood inference. For the evolutionary model, I chose ‘Rapid bootstrapping’ and run the it. Once the process was finished, I built a consensus tree by selecting the ‘RAxML bootstrapping trees.’  I found that the clades with high support do not match those from your bayesian analysis.

Next, we used a different program that uses maximum likelihood called ‘PHYML.’ We ran this analysis with bootstrapping and we used the HKY85 model of molecular evolution for this and our final Bayesian Tree.

Results from our ‘In Lab’ portion:


At Home portion:

Our Assignment for the second portion of this lab was to use the same alignment and the HKY85 model of molecular evolution to infer my best tree using Mr.Bayes. For this run the only modified setting were the ‘chain-length’ was set to 3,000,000, the ‘burn-in length’ at 300,000, the ‘subsample frequency’ at 500. This run lasted for about 5 hours long!

Result from our ‘At Home’ portion:


Lab 6: An Introduction to Geneious

October 1, 2019

We began our lab session by downloading a software called Geneious. Geneious is a commercial program that performs many functions on DNA and protein sequence data. I first created a folder where I was able to download and store Fish DNA barcoding sequences from canvas. The folder contained both the reverse and forward reads and contained the chromatograms of our sequences. However, when I clicked on the forward read of my sequence it looks like I did not have a very successful PCR run at all. This may mean that bacteria or fungi could have contaminated my sample or that my sequencing run was a complete fail. Because of this, I used the sample sequences from other students which were much more successful in their PCR runs. I used the files “ARA01” and “BP01” as my new sequences. After I dropped these files into my “Fish Barcode” folder, I was able to begin assembling the forward and reverse sequences of the sample “BP01”. In this step, I was able to examine my sequences and delete any bases on the two ends that were unreadable or trim off any ambiguities. Once that was completed, I generated a consensus sequence and put it into a new file where it was ready to undergo BLAST. BLAST stands for Basic Local Alignment Search Tool and is used to search the database for highly similar sequences. The top hit for my sequence BP01 was Thunnus obesus which is yellowfin tuna. This is exactly the name of the fish that the student was told when they purchased it. For the sequence ARA03, I found that the top hit was Seriola quinqueradiata which is Yellowtail or Japanese amberjack. Once again, this is exactly the name of the fish that the student was told when they purchased it. There were polymorphisms present in both of these sequences. In the ARA03 sequence, I found 599 polymorphisms which were all SNPs. The first 10 polymorphisms were at sites 63, 68, 70, 89, 92, 95,113,128, 131, and 146. In the BP01 sequence, I found 8 polymorphisms which were also all SNPs. The polymorphisms were at sites 9, 275, 291, 363, 384, 405, and 418.


Lab 5: Finding Mimulus

September 24, 2019

The first stop on our field trip was Mount Tamalpais. When we first pulled into this area, we saw that there was a natural stream of water coming from the mountain. Here we were able to observe a large costal parenial Mimulus. It was important to note that, we found these mats or patches of Mimulus directly next to the water source. Mimulus loves to keep its feet wet so it is not surprising that we found Mimulus near a spring. In a location like Mount Tamalpais, we were able to see how important the bee population is. Usually, the bees are the ones to pollinate the flowers along this mountain side, including Mimulus, which promotes intrabreeding and interbreeding amongst plant species. We saw that the baby Mimulus did not have much space to “move” or pollinate on its own so, it is vital that bees are able to pollinate around the mountain.

Professor Paul brought it to our attention that the Mimulus population we saw could be a sink population. This means that the population would thrive near a body of water but may get flooded out by rainfall before they are able to flower. Additionally, Alec shared that in drought years, Mimulus may thrive better because they have a longer time to flower and germinate their seeds. This would be beneficial because these Mimulus would be more likely to contribute to the next generation. I was able to learn that Mimulus responds to long day periods but this time of year, there are shorter day periods so, it may not be long enough for these Mimulus to get the proper environmental cues to grow and persist.

We then drove a few miles away from the mountain to a location called Redwood Creek. The creek was well shaded and made for a cool and moist environment with patches of sunlight for Mimulus. Our task in this location was to find a Mimulus population that was growing along this creek. The class as a whole was not able to locate these populations but, we managed to find a few small Mimulus with the help of a few careful eyes. At Redwood creek, it seemed that the populations of Mimulus that we found were much smaller and hidden than the past populations that we had seen. It was explained to us that plant species that live along Redwood creek are susceptible to flooding during the winter time. This, again, is detrimental to Mimulus because they are not able to properly germinate and contribute to future generations. Since the creek stretches for miles, it is apparent that the separated populations of Mimulus have been limited on insect pollinators that are able to travel long distances to pollinate.