Author: Prof. Andresen

Today I finished preparing my new calibration standards, as well as preparing another sample set. I am running the calibration standards for both the calibration and as a sample, so this should help me investigate how the calibration process is working. The new sample set should help to see if the calibration is helping to reduce the error. I am also trying to use a nonlinear fit for the calibration, to see if that helps. I am in the middle of analyzing the data now, but it is unclear whether I have been successful yet or not.

Yesterday we went to a pretty crazy farm to get our chicken blood. The experience was crazy what with the way they killed the chickens, the machine they used to get the feathers off, and the chanting women in a factory line plucking the rest of the feathers. After we got back to campus we began by separating the blood evenly to seven tubes and then putting them through a centrifuge at 2000 times g. after ten minutes we took the samples and removed the plasma, the yellow liquid at the top that had been separated from the blood. We then carefully removed the white coat that sat on top of the red blood cells. This was a lengthy process because it involved us washing the blood with a solution the professor made, running in the centrifuge, and then carefully removing the wash and whatever white stuff we could. After doing this three times we were finally down to just red blood cells and we put those into new tubes and froze them at -70 c.

Yesterday, I spent the morning analyzing the results from the previous day’s test. The results were somewhat inconclusive, but they did give me one useful piece of information. In general, when I used the 1000uL pipette twice, the results were slightly more accurate, by around a percent. The results were still off by approximately 3-4% though, and even the samples that I set to match the calibrations standards were off by this percentage. I tried altering the calibration method, which lowered the uncertainty a bit, by around 1-2%, so I believe the error is coming from both sources. In the afternoon, I designed and began preparing a new set of calibration standards, which go over a larger range and with more points in between. This should help me to minimize the uncertainties.

This morning I finished the RCR training modules, and we also had a training session on chemical safety. During the afternoon, I prepared and tested my two new magnesium series. Each series has 9 samples, ranging from 0mM Mg to 4.11 mM Mg. One series was diluted using the 1000uL pipette, and the other was diluted using the 5000uL pipette. Also, several of the data points should match up with the calibration standards that we use. This should help me identify if the uncertainty is coming from the solution preparation or from the machine’s calibration. Tomorrow I will be able to analyze this data and possibly design a new test if necessary.

Today was another relaxed day I studied up some more on the nucleosome and the histome octamer which is what the DNA wraps around. I learned that as well as the hockey puck shape that the histome octamer can have there is also a sort of V shape with two spheres at the end. I also tried to set up a group file between the three of us so we could post our work and be able to access it anywhere but the gettysburg system got the better of me and we will have to have IT set that up for us too. Tomorrow we are going to go get the chicken blood and then create our nucleosome samples.

Yesterday in the morning, I did a bit more research into how the plasma torch works, as well as the rest of the spectrometer. I gave my presentation on the work that we will be doing, and I also listened to what everyone else is up to. Then, in the afternoon, I worked on preparing a new set of samples to test in the spectrometer. This set should help me understand where some of the uncertainty is coming from. I am going to run two more series, which include some points from the calibrations. This should give us some idea if the calibration method is affecting the results. I am also going to prepare the samples using two different pipette methods, which will test another variable. Hopefully this will give us some idea where the uncertainty is coming from.

This morning, I analyzed the results from my tests yesterday, and I found some strange results. In general, all of the predicted solution concentrations I got from the spectrometer read-outs were at least 3% larger than the concentration to which I tried to prepare the solutions. In fact, all but one set of data was over the prepared concentration. This is rather a rather disturbing systematic error, and the sources need to be identified. Though the results are not as accurate as we need, the good news is they are very precise, because if you ignore their consistant 3-6% overreading, they are then within 1%. These could be all random errors, but that seems unlikely because of their consistency. They could also being coming from my technique of solution. It could also be a systematic error coming from the pipettes. It could also be an issue with the calibration with the machine. I will investigate the source of the error more tomorrow. I did a few tests at the end of the day, and they showed that the pipettes might be cause a small portion of the error, but nowhere near all of it. I also did more research into the cobalthexamine project that I will be working on throughout the summer.

Today I increased the amount of MgCl I added to the solution in order the speed up my result process. For a few trials my solutions stayed at or below the .1 concentration mark but as time progressed my concentrations began to increase at around 48 micro liters of MgCl being added to the DNA solution. This meant that the nucleosomes that had clumped together earlier and fallen to the bottom of the solution were breaking apart and floating back up to the top. I never got my results up to a definite high concentration because as I kept taking data the Professor noticed that the amount of absorption at the 280 nm mark was very high and getting higher which according to him meant that there were excess proteins in the solution which could be the result of bacteria growing in the solution. I think the problem might have been that I started out with very small amounts at the beginning which slowed my result process and caused it to be a lengthy experiment. Also the solution that i was using were leftovers and might have just been too old. I hope to try and redo this experiment after we create our new nucleosome solutions in the following weeks and plan to use my results from this trial to influence how I go about taking data with the next experiment.

I spent the first half of the day today organizing and analyzing the results from yesterday’s samples. The results I received for the predicted concentrations of my soultions were pretty consistently 5% away from the values that I expected. This may be a problem with my preparation of the solutions, or it could be an issue with the calibration or conditions of the machine. In particular, the axial readings for the Magnesium concentrations were consistently 30%+ away from the expected value. For the most part, these results should be ignored because they are oversampled. Also, I found that consistently the first sample I tried to run was not drawn from the test tube, which led to a lot of strange negative results for one of the samples.

To try an look further into these issues, I prepared and tested a large new set of samples. This new set has two magnesium series, a cobalt series, and a phosphorus series. This should help identify where some of the 5% error is coming from. If it is a problem with the machine or the range of concentrations we are running, more data will make this easier to see. If it is a problem with my preparation, this will also become obvious. I ran the tests at the end of the day yesterday, and will analyze the data tomorrow.

Also, we are looking at changing the calibration method for the tests, because of the strange magnesium readings we were getting for the axial view. It is possible to use a non-linear fit for the calibrations, and we will see if this helps tomorrow.

Today was a continuation of yesterdays research I continued to add MgCl to my DNA solution and test the amount of absorption of the light we put through it. After hours of taking data i finally got to the bottom of the curve where the MgCl has pulled most of the nucleosomes to the bottom of the solution and have just started to get results of the nucleosomes freeing themselves and returning to the top of the solution. I will continue to add larger amounts of MgCl to the solution tomorrow and hopefully get the data that I am looking for.