Tag Archives: recap

D2O Adaptation Day 8

I setup new cultures for 99% D2O and 50% D2O making this the 4th generation of yeast in D2O. Yesterday I ran out of D2O YPD and had to deal with a smaller batch for the 50% D2O yeast. Today I made some new 99% D2O YPD and inoculated 1ml of each yeast-water type in 9ml of each YPD type. Then I took some measurements.

Today’s Absorption measurements are:

  • D2O (Generation 3) – 2.351 at 24h
  • 50% D2O (gen 3) – 2.65 at 24 h
  • D2O (gen 4) – 0.041 at 0h
  • 50% D2O (gen 4) – 0.419 at 0h

It was interesting to find that the 24h batch of 99% D2O yeast had an absorbance of 2.351 for 2 reasons; 1) that’s a good amount for 24h from what I’ve been observing and 2) because my 1:10 dilution resulted in a much different absorption than 50% D2O yeast (note that they started at a similar value).

When I tried to analyze by eye it was apparent there was a difference in absorbance but it definitely didn’t look 10x less clear. But I did notice that the cells in 50% D2O seemed to be pretty uniformly “dissolved” (I don’t really know what to call that). Meanwhile, the cells in 99% D2O seemed to form larger clumps and the average “particle” size (I’m guessing they are colony clumps) was larger than the clumps in 50% D2O.

I’ll have to analyze this under a microscope and report observations, something that will have to wait until tomorrow or next week.

 

SDM ligation results

So the good news is that had I not thoroughly messed up the reaction, this would have been a resounding success. The other good news is that this part of the experiment is fairly simple to repeat using the pRL574 anchor. The bad news is the ligation went as I expected it to which is to say sloppily.

Now I have to try to explain to you these results:
image
Above is an image of the gel under UV light stained with Ethidium Bromide. I used to use Sybr Safe for visualization (and have in the first part of these experiments), but in the past unzipping was never successful so we think there is a chance that Sybr Safe interacts with DNA differently than EtBr and could hinder unzipping. So in an effort to identically repeat Koch’s grad school work I’m using EtBr because that is what he would have done.

Below is a cropped and grayscale version of the above image:
image

We are looking at a lot of lines here. There is a meaning to them all though:

  • The two outer lanes are 1kb ladders in different amounts. More specifically the right most lane has twice as much ladder as the left most lane. We’ll call these lanes 1 (left) and 5 (right).
  • The three middle lanes from left are: ligation with 5′-bio adapter (lane 2) and pRL anchor, internal-bio adapter and pRL anchor (lane 3), and 5′-bio adapter with pALS anchor (lane 4).
  • The ladder reads (from the bottom): 1kb, 1.5kb, 2kb, 3 (brightest band in the ladder), 4, 5, 6, 8, 10kb
  •  Lanes 2 and 3 are identical length wise and the only difference between the molecules is where the biotin molecule is located.
    • The band at the very bottom is the pRL574 PCR product band which is 1.1kb in length.
    • The next two bands above that are pBR322 that has religated into circular DNA.
    • The band above that is digested pBR322 (4.3kb).
    • And the band above that is the unzipping DNA that we want, which is around 5.4kb. I have gel extracted that piece.
    • All bands above this band are chains of pBR322 that ligated together and may or may not have the anchor and adapter ligated to it.
  • Lane 4 is interesting. I would be inclined to say that we got better ligation out of this band because there is less circular DNA bands (compare with lanes 2 and 3). But the product is harder to quantify.
    • The brightest band is the pALS anchor, 4kb in length.
    • Barely above that is the digested pBR DNA, 4.3kb in length. And is identically positioned with the other two lanes.
    • Above that is a very feint band and now I’m inclined to believe that this band represents two pBR322 fragments ligated together and circularized. (Above I say its the product of the reaction that I wanted). I think this because this band lines up exactly with bands in the other two lanes.
    • The next visible band I believe is the product that I expect, 4kb pALS + 4.3kb pBR -> 8.3kb unzipping sequence. It is right there, but again bands in the other lanes could indicate something else.
In orange are the bands that I gel extracted using a razor and incredible wit.

Well tomorrow I’ll gel extract and hopefully we can get some DNA tethers by the end of the week and try for unzipping. Also tomorrow I’ll retry the ligation reaction for pRL-pBR (since I have no more predigested pALS). Tethering is the only real way to identify successful ligation so we’ll find out together!

PCR machine testing: results

Via figshare:

OpenPCR and Thermo PCR Sprint Thermal Cyclers testing. Anthony Salvagno. figshare.
Retrieved 22:53, Aug 22, 2012 (GMT)
http://dx.doi.org/10.6084/m9.figshare.94408

Well, well, well. It looks like OpenPCR works much better than my ThermoCycler (which I just learned today is actually called PCR Sprint). The temperatures match pretty well between what the program is set for and what the recorded values are. I’ve discovered in the past that OpenPCR has a problem getting above 90C (see bottom of post), but I don’t think that is an issue here. If OpenPCR isn’t producing product then I may not have chosen the right annealing/extending temps or the problem may be elsewhere (unlikely).

ThermoCycler on the other hand is a piece of garbage. I’ve always hated this thing and now I have proof (again, I’ve done an experiment like this in the past, but it wasn’t as bad as these results show). Looking at the file “thermo-oil-3-cycles.png” you can see that ThermoCycler never gets below ~62C which is sad because it is set to anneal at 52C. It also extends at ~74C when it should be near 69C.

Tomorrow I’ll be taking some T measurements to try and get the program to be closer to the temps I want to run the pALS protocol at. OpenPCR will just need some slight modifications, but ThermoCycler needs a lot of help. Sigh…

PALS PCR 4 results

image

  • 0.8% gel: 50mL of 1x TAE, 0.4g agarose, prestained with Sybr Safe
  • out of 10 reactions only 1 reaction worked… hmmm. And there is a chance that the product isn’t even what I want. It appears that the band is not 4kb, but the gel didn’t run evenly so there is a chance it is 4kb but doesn’t appear that way because of how the current pulled the DNA.
  • It looks like a machine test is in order. The Thermo cycler said the program completed today (which it hadn’t in the past few trials), but also said the program completed in 2.5 hours (which it shouldn’t because it is a 4.5 hour program). And the OpenPCR won’t get below 16C for the final hold even when I set it at 15C, which it should not have a problem holding at (I’ve seen it consistently get to 12C).
  • I will also order new dNTPs even though I think that may not be the problem.
  • I also learned yesterday that the freezer wasn’t closing completely and everything may have thawed potentially ruining the enzyme.
  • Finally I should try to mix the master mix better if that is an issue at all.

Lots to do in so little time… Gotta get this working by this weekend.

Open notebook science thoughts inspired by the biomedical research symposium

First let me say that my presentation went amazingly. I spoke for ten minutes with a five minute Q&A session after. The audience was comprised of IMSD students, their faculty advisors, and their friends and family. It was also easily the largest audience I’ve spoken in front of. I estimate about 70-100 people.

The Q&A was quite exciting. I received questions from every audience group: students, faculty, and nontechnical attendees. This is very important to me because it tells me that my presentation was engaging and impactful. The questions used the entire allotted time and it spilled over into break period.

During the break I spoke with students interested in open notebook science, participating in it, and reaching out to others who could be interested in it. I also spoke with nontechnical audience members (the friends and family) about the core values of ONS. This was very exciting to me because it shows me that taxpayers and the general public care much more about science and information more than we give them credit for. I always say ONS is more than just data provenance, but is also scientific outreach, and the conversations I had today demonstrate that.

I also received some interesting comments from faculty. I always receive comments from this crowd regarding two things: 1) information signal to noise and 2) data protection. Here are my thoughts regarding these two issues:

1) Scientists already publish a ton, and there is already a decent amount of bad science. Open notebook science does not add to this perceived signal to noise ratio. In this case signal to noise refers to the amount of relevant scientific information (for your interests) in the sea of all publications. By providing open access to your data and keeping a complete account of your research you are actually making it easier to find the information you want and need.

If you are reading a journal article, you have to sift through the document and spend time trying to understand what the author is proposing with their conclusions. With open notebooks that information is laid out for you in plain english with minimal effort required. You need a protocol? Here are my steps. You want the raw data? Here you go.

While it is true that open notebooks will increase the amount of published scientific information in the world, in this case that overload enhances the discovery process and minimizes the time required to follow up on prior experiments. I can’t count the number of times I’ve saved oodles of time by finding information in my notebook and the informally published documentation of others.

2)  When it comes to data protection, I will admit that I don’t know everything there is to know about this. I also have never been a victim of data theft (scooping) myself. But I don’t see how open notebooks can increase the frequency of data thievery.

Firstly, bad scientists are trained to be that way. There aren’t many people that want to be a bad or ethically incompetent scientist. Those that are trained to be that way. In graduate school you are prepared to lead the next generation of science and you learn from those around you. If your PI follows a negative code of ethics you wont have all the necessary tools for success upon graduation. Luckily even those who are raised in this environment, like my advisor Steve, have the choice to follow those guides or choose a different path. In the case of Steve, he was so uncomfortable in his environment he chose to be an open scientist. This decision ultimately led me to becoming an open notebook science evangelist, a career path that has led me on a wonderful adventure.

Aside: I really hope I can convince Steve to air his grievances publicly because his experiences are something that all scientists can learn from.

Secondly, data thievery already occurs in a closed environment. I’m sure someone one day will be a victim of these circumstances in an open environment (if it hasn’t already happened), but being open won’t open the flood gates on scooping and unethical science. In fact, I believe that open science can help minimize it.

By being open you are essentially prepublishing and staling claim on your research domain. In the event of catastrophe, you can point to the fact that you have been working on this project and that your information is valuable and full of integrity (integrous?). Peers may also be able to back you up and essentially police the situation accordingly. Being open makes your research transparent and could help prevent tragedy. Why would you choose to steal information that others already know exist?

And to that point, why would you choose to steal information that is encouraging reuse? By participating in open science and open notebook science and publishing your data with open access you are encouraging data sharing and reuse. You can’t steal information that is being given away.

I think the issue is that scientists who work in the closed environment think their research is theirs and the protect it like it is an extension of themselves. Open scientists, on the other hand, view their data as that to be shared with the world. Their data is not theirs but something that they produced and should be consumed and developed by others in ways they can’t imagine. Traditional scientists who think about open science continue to view their data as theirs in the open environment. The truth is the two systems are fundamentally different and require a mental reprogramming in order to go from closed to open science.

As I always say, education is the required mental reprogramming and when this happens we will see a much faster shift from one system to the other. I envision a world that is based on the quality of research you produce, not the quantity or the perceived impact of that research.

Personally speaking, I feel that younger scientists can embrace open science much better than the older more established generation. The older generation was scientifically raised in a system that embraced a different set of core values. New scientists haven’t been influenced as much by the current system and are more open to change because of this lack of influence.

With regard to data theft, I want to add that I wouldn’t mind being scooped. This may sound strange to invite unethical conduct, but how can anyone understand the negative aspects of the system without experiencing it themselves. Unfortunately all my data is public domain so data theft is essentially impossible. Perhaps I don’t get a credit or a citation if someone reuses my data, and that would basically be the most unethical thing that could happen to me. And if that is in fact the worst case scenario then I would say the system works pretty well.

pALS PCR Results OpenPCR vs Thermal Cycler

Yesterday I setup a PCR reaction using both our labs thermal cycler from Thermo and our OpenPCR thermal cycler. I want to make pALS PCR fragments (4kb in length). And here are the results of that experiment (with setup):

Gel Setup:

  1. 50ml 1x TAE buffer mixed with 0.4g High Resolution DNA agarose
  2. heated in microwave for 2 min
  3. added ethidium bromide (EtBr) to final concentration of 1ug/ml
  4. gel cooled for 40 min in fridge
  5. Meanwhile, 5ul of 9 of 11 PCR reactions put in PCR tubes with 1ul of 6x loading dye added (for 6ul total per tube)
    1. the tubes selected were 4 tubes from OpenPCR (numbered 1-5) and 5 (of 6) tubes from Thermo thermal cycler (numbered 6-11), so tubes 1-4 and 6-10 were selected for gel analysis. This is because there are only 10 wells and 1 well is needed for DNA ladder.
  6. Once gel cooled, 250ml of 1x TAE added to electrophoresis device (what’s the name of this thing?), filled to cover gel
  7. 6ul of each of the 9 prepared tubes were placed in the wells along with 6ul of pre-prepared 1kb DNA ladder.
  8. connected to power supply and run at 150V for 45min

Results:

image
Gel results. Lanes 2 and 4 have feint bands where the expected PCR result should be.
image
Enhanced gel results.

Based on the images above, the PCR reaction was a failure. Lanes 2 and 4 have very feint bands where the 4kb product should be, indicating the reaction succeeded but not to the degree required. This could be for any number of reasons, but most likely due to inactive enzyme, old reaction buffer, degraded dNTP’s, etc. Basically I’ll be doing this reaction again next week when all my new supplies come in.

Notes about the image acquisition:

Since I ran the gel with the EtBr added to the molten gel, I did not need to run the gel and then add EtBr. I also did not destain the gel after completion of electrophoresis. The images above were captured with my phone camera. To see gel results typically I use SybrSafe from Invitrogen with their special illuminator, but since I used ethidium today I had to use the hand held UV lamp we have. This means worse quality photo, but oh well.

An alternative analysis to FT-IR to study deuterium exchange

Via figshare:

Deuterium Content of Deuterium Depleted Water: 1st Trial. Anthony Salvagno, Scott Jasechko. figshare.
Retrieved 21:24, Jul 20, 2012 (GMT)
http://dx.doi.org/10.6084/m9.figshare.93089

A few days ago I made a new friend named Scott Jasechko. We met to discuss the possibility of creating a TED branded forum to UNM. We got to talking about our research and it turned out that our research interests are very much aligned and he studies water isotope amounts with relation to natural water around the planet. He’s about to publish his findings and once that goes through I’ll link to that.

It turned out that his lab has a device that can very accurately measure small concentrations of deuterium and oxygen-18 in water samples. I told him about my DDW experiments and how deuterium exchange may affect my experiments, but that I can’t quantitatively measure it’s affect or the process in general. So we got to talking and he wanted to help me out.

The data linked above is the results of the mini-collaboration that I predict will turn into more. Scott used his Picarro cavity ring down spectroscope, which means very little to me right now, to analyze the water samples I gave him. The water was used in these two experiments (each word is a separate link). And was then stored in our desiccator (with drierite to reduce moisture exchange) until yesterday (July 19, 2012).

Surprisingly, the data shows very little change from what Sigma claims (less than 1 part per million D to H) to yesterday, showing very little exchange. There are two things to consider here: (1) the machine wasn’t calibrated for such low levels of D, which skew the readings (since they actually give us negative ratios), and (2) volume may play a part in exchange.

I don’t understand the mechanism very well but I suspect that surface interactions play the largest role in deuterium exchange. Once deuterium is introduced into the sample, then I would guess diffusion takes over, but this is probably slower in nature than evaporation/other mechanisms that are involved on the surface. Basically this is a thermodynamics problem that I would need to spend 3 months thinking about to compare with experimental analysis (since that is how long the Thermo class is, 1 semester).

The follow up to this experiment should be better organized. Obviously we’ll need to redo this experiment. Scott and I are also performing another experiment where Scott has left the lid off the samples I gave him so he can see if there is a new value the spectroscope provides. From here we may want to do some longer time analysis and some other studies that we’ll have to plan. I’ll start a new thread for that, in the mean-time I hope he’ll introduce himself in the comments of this post.

Some other notes:

  • I geeked out when I realized the power of this sort of collaboration. If I hadn’t been affiliated with TEDxABQ and thus this new idea of TEDxUNM (not officially licensed), I would have never met Scott. These are the sorts of collaborations that I hope can be introduced because of the Open Research IGERT proposal.
  • I mega-geeked out when I realized I can do a project planning thread with someone here at the university in another lab that can also participate in the project! Open science at the core all the way.
  • I super-mega-geeked out when I realized that I implemented a crucial aspect of ONS. As I wrote this post I wondered what experiments I had used the water for. Then I realized that it’s all documented and that I can show everyone what those experiments are. It’s not enough that I have it recorded by date, but also that you all saw those experiments in real-time and their use in future experiments has been realized. That’s a major win for ONS.

Yeast Time Trials in DI, DDW, and D2O: Trial 5 Results

Via figshare:

Yeast Hourly Growth in DI, DDW, and 99% D2O. Anthony Salvagno. figshare.
Retrieved 21:15, Jul 03, 2012 (GMT)
http://dx.doi.org/10.6084/m9.figshare.92771

Notes:

  • I had a meeting this morning regarding the IGERT I speak so frequently of. That means the timed data wasn’t so consistent in the beginning. I took my first time point about 25 minutes after I set up the experiment. And then I took my second time point 2 hours later. The raw data will show that I took the first time point about 20 minutes later, but I assure you the time difference was more than that.
  • I did take the rest of the time points an hour apart.
  • I spilled a considerable amount of the D2O sample (which was in the beaker), maybe like 30-40% of it after I took the first time point.
  • In data news, I find it really strange that the yeast in D2O didn’t grow nearly as much as it did in the past. I’m wondering if the new setup had anything to do with that.
  • Also it looks like yeast in DDW grew more than the yeast in DI. Again I’ll have to do another run to verify, but I feel like this could be a real result.
  • The added volume definitely improved the results, as the yeast remained suspended in the flasks. In the previous trials I would get considerable settling because the test tubes wouldn’t swirl as much. Much better setup.

So the moral of this story is, I’ll have to do another trial of this setup. Fine by me! Whatever it takes to get good results that are repeatable. That’s the nature of open notebook science!

Yeast hourly growth in DI, DDW, and 30%, 60%, 90%, and 99% D2O

Results:

Yeast growth in DI, DDW, and 30%, 60%, 90%, and 99% D2O. Anthony Salvagno. Figshare.
Retrieved 21:20, May 24, 2012 (GMT)
hdl.handle.net/10779/caa57855586213d79eea5576f8da23d0

Notes:

  • I don’t think this is a reliable data set. Several of the samples actually read a lower absorbance after the first hour than they initially do. And then they all dip again at hour 4.
  • I noticed a considerable amount of cell settling after hour 3 on the bottom of each test tube. I mixed prior to reading the absorbance values, but this is likely to skew the results. Next trial I will have to mix before reading every hour.
  • The data between DI, 30%, 60%, and 99% D2O look consistent with Tuesday’s results, but it scares me that the DDW and 90% D2O are completely out of whack.

E. coli growth in different water types data

Via figshare:
E. Coli Growth in DI, DDW, D2O, 30% D2O, and 60% D2O. Anthony Salvagno. Figshare.
Retrieved 23:35, Apr 27, 2012 (GMT)
hdl.handle.net/10779/51fcd2f94fd7464449ee0f794642214c

I’ll post some interpretations here later…