NLP week 4, 5: a crazy (but lazy) workaround and some machine learning

Hi!  My name is Jiehan Zheng and I work at CCLS at Columbia University on natural language processing and machine learning.  I've been doing some training data collection, evaluation and postprocessing work in previous weeks, and finally in week 4 and 5 I get to do some real machine learning!  It's been busy but interesting two weeks.  I did too many things and I am not sure if I can recall all of them...but let me try--

After building the model comparison framework in HTML and JavaScript, Apoorv asked for a new feature--calculating p-values from Χ^2 from McNemar's test for models to indicate how differently any two models perform on the dataset.  I had no clue how to do this, even after looking at Wikipedia and several papers, so I asked Apoorv how he used to calculate the p-value before having my framework.  He sent me a MATLAB function file that he used to use at IBM.

I tried the MATLAB on Columbia server and verified that this function file works.  Then another software, Octave, immediately came up to my mind.  Octave is open-source and although it doesn't advertise as so, it is known to be an open-source "implementation" of MATLAB's features, so that everyone can use it freely.  So I ran the function in Octave as well and it works too.

I then looked into the source code of that MATLAB function (although I've never used MATLAB before...), and found out that although most of the calculation steps are fairly simple and straightforward, it calls a function called chi2cdf() from MATLAB.  By looking at MATLAB documentation I found that the chi2cdf() function, as expected, contained a definite integral in it.  So I basically ran into the same problem of not being able to calculate definite integral in JavaScript.

Then, I don't know why, a crazy (but lazy) idea came to my mind...  If it is hard to calculate definite integral in JavaScript, then why bother doing it in JavaScript?!?!  I can simply install Octave on my server and set up an API on the server to pass the X^2 to Octave and ask Octave to calculate p-value for us!  I quickly installed Octave and wrote up a short Node.js program that listens for requests, and spawns a new Octave process whenever it receives a request, pass the X^2 value to Octave's chi2cdf() function and collects its output then returns it to the browser.  So whenever Apoorv enters the command to calculate p-value in the browser, my code is going to send a request to my server and wait for the server's response and display that answer.  And very luckily I proved this idea actually works and was eventually able to implement this in a few hours.

Well, that's all I can say for now...which is the work I did in the afternoon on July 4.  For the rest of the time I wrote code to extract features and making training examples (without any annotation from human!) in preparation for a machine learning task (sequence labeling), and got some training data from several websites and stored them to a MongoDB database.  I also posted an answer on Stack Overflow for the first time while looking for solutions to a PyMongo error and after reading some MongoDB's documentation!  Unfortunately I can't share more details on the machine learning task for now but I will in the future!

Oh, and on July 4th Apoorv asked me if I'd like to see fireworks at Dr. Rambow's apartment building.  I went and it was very amazing!  I also changed my plane tickets so that I can extend my internship by a few days.  So now officially I will work at CCLS with Apoorv for almost 7 weeks!

Thanks for reading!  See you next week!

The Gab Lab: Week 5/Last Week

Hi everyone. This is Michelle again, and I'll be posting about my fifth and final week at the Gabrieli Lab at MIT. This week was a little different in terms of less data entry work and more sort of 'one-off' opportunities. On Monday, I spent the morning as usual: updating current subject info on the database, scoring, and recording. However, I spent the afternoon observing an EEG, which meant that I was unable to attend a lunch meeting with Dr. Aaron Matfeld. Each EEG session is roughly 4 hours, as there is a lot of prep work and clean up involved. Nayeon and Calvin first showed me how to cleanse all the electrodes by soaking them for 10 minutes in an ionized solution. Then, we had to gel each individual node, and each node's corresponding attachment slot on the EEG cap.  When the subject came in, we adjusted the EEG cap on her, attached the chin straps, and in addition to plugging in all the nodes, we also attached special nodes under her eyes and ears to measure the vertical and lateral eye movement that I posted about several weeks ago. 

EEG Nodes

EEG Cap

When everything was set, we placed the subject in a dark room and monitored her brain's electrical signals as she performed various language training tasks such as the syntactic/semantic error test that elicits the N400 or P600 response. Next, we ran the first MAL training session with the subject. Since I've mentioned scoring and testing so much in all my posts, here's a quick rundown on how these testing sessions work. Each session consists of 30 vocab questions and 30 grammar questions. During the vocab section, subjects see a picture, and have to say the correct word in MAL. In the grammar section, subjects see a scene, and have to produce a grammatically correct sentence that describes it. A correct sentence has this layout: subject - article - object - article - verb - article. However, there are two classes of subjects and objects, and each class uses a different article (Ihn or Ihd). Furthermore, the verb's article (Ahn or Niy) is entirely dependent on the subject's class.

Example of vocab testing

On Tuesday, I met up with Kelly, the lab's psycho-educational evaluator, to talk about the psychological context of the tests I have been scoring. We talked about the KBIT (IQ test), the CVLT (you hear a list of 16 words and have to recall them over various periods of time/in various ways), the CnREP (you listen to made up words and have to reproduce them perfectly), and the WJ3 Sound blending test. Kelly ran a little bit of the sound blending on me, and I have to say, it was surprisingly difficult. In this test, you have to reproduce a word after hearing a recording of it that has been split up phonetically. For example, the word 'grasshopper' sounded like 'guh-ra-ss-huh-opp-er'. Kelly also talked about the role standard deviation, standardized scores, and normalization play in these tests. 

On Wednesday, Cirkine and I got to shadow Maiya in the hospital! In addition to researching at the Gab Lab, Maiya is a neurologist at Brigham and Women's Hospital in Boston. Maiya showed us around the hospital, and we sat in on an interdisciplinary meeting. During this meeting, which is attended by 30 or so psychologists, neurologists, and neuropsychologists, several of these doctors spend 10 minutes presenting puzzling cases, ranging from memory loss to language problems, that they cannot seem to diagnose or solve. They show brain images of these patients, give us a brief medical history, and so on. Then, everyone in the room asks questions or gives suggestions as to what do next. I wish I could talk more about these cases as they were truly very fascinating, but I had to sign a confidentiality agreement. Because I was away for the day, I missed the weekly stats meeting with the LAP/CASL members, but Calvin was nice enough to send me a video that covered analysis of variance (ANOVA - basically a t test for multiple samples) to make sure that I was up to date.

The next morning, I did more sentence production scoring. Then I had my last meeting with Amy, in which she gave me lots of data for my poster. She also suggested that I send her multiple drafts of my poster so she can give me feedback; Thanks Amy! Next, all the interns went down to the MEG (Magnetoencephalograpy) lab for a tour (organized by Joanna & led by Demetrius). MEG is basically the same thing as the EEG, but it looks completely different, and uses magnetic fields instead of electrical signals to look at the brain. 

MEG (Your head goes up that tube thing!)

The MEG machine is in its own small room. Demetrius told us that the room weighs over 7 tons and is made of 3 layers of metal alloys in order to prevent magnetic fields from interfering with the process. 

Rather fittingly, my last day today was extremely relaxed and low-key. I spent the entire day scoring sentence production and saying my goodbyes/thank yous. 

Thank you to John Gabrieli for allowing me to participate in his lab. Thanks to Amy, Zhenghan, Michelle, Calvin, Jenni, Nayeon, and everyone else for an amazing experience. Thanks to Dr. Peretz for the opportunity to work in a lab and preparing all of us properly. I can say that after just 5 weeks at the Gab lab, I had a lot more fun than I thought I would, had a fantastic time, and would really love to pursue more research in the future. 

The Real Deal Week 4-5

After a few weeks of waiting and disappointments, we're finally on a roll again. The last few weeks have been slow because we had a major contamination problem... meaning the plates we use to grow our worms had other bacteria and/or fungi growing in them. It took a while to pinpoint the source but after making batches of fresh plates, flasks of fresh E. coli and after taking extra precaution we finally have a stock of contamination free plates. However that was only the beginning of what’s to come…

The practice lifespan I started during my first few weeks in the lab is now coming to an end and the results show that the mutant worm is living much longer than it should be. The gas-1 strain (mutant strain) is supposed to have a median lifespan of about 8-10 days but my lifespan experiment shows that they are living 18-20 days…almost as long as the wild type (normal strain). Though I had begun prepping for my actual lifespan using drugs, my PI said that no experiments could be conducted with an animal that has lost its main phenotype. What she means is that the worm’s life span is a major phenotype that we use for experimentation and if this phenotype is not apparent in the animal we’re working with, then any experiments we do using this animal are worthless. All of the experiments in the lab that have  been working on has also been affected by this discovery. The mutant worm could be living longer than usual for a few reasons. 1. A random beneficial mutation may have occurred in one of the worms and then outlived the mutant worms and reproduced (in a sense survival of the fittest)

2. A wild type worm may have somehow crawled onto the mutant worm plate and then outlived the mutant worms (again survival of the fittest)

So long story short, my original drug lifespan has been put on hold until we find a solution. 

Tissue Morphodynamics Lab: Week 6

So this is final addition to my three part post. The sixth week was just as exciting and new as the fourth week, as I was given a new project to work on for the rest of my stay. Needless to say, I had to do a little bit more reading of reviews and articles to understand the basics of cell mechanobiology, but it was all very similar to the reading I did the previous summer at a Head and Neck Cancer Lab at UCLA. The post-doc, Jason P. Gleghorn, and the graduate student, Sriram Manivannan, explained to me their hypothesis, their prediction, and their results so far regarding cellular motility and cellular divisions along an axis. It was essentially the same topic I listened in a group meeting two weeks before.

The specific aim of their research is to understand the mechanics behind cellular division and how rotational axes and endogenous stress affects the division of cells.

In addition to culturing, dissecting, fixing, imaging, and analyzing chicken lungs, I was given the task to analyze their image data of hundreds of 250 micron by 250 micron squares of cells over a 12 hours period in 6 minute intervals. There are 121 images per time lapse that are compile into a TIF file and then analyzed using Imaris, a cell tracking program. Then using MatLab, the data of all the positions of the parent and daughter cells are compiled into an image that shows the overall positions of all the time lapses.

Imaris program. Tracking the parent cell. 

Imaris program, Tracking the two daughter cells. 

Jason usually deals with compiling the data, whereas Sriram used to analyze the images as well as create 2D patterns, culture the cells, and produce the images on the confocal microscope over the weekends. Now that I'm analyzing the majority of the images, he has more time to write up his paper regarding cell motility as well as prepare for group meetings and conferences.

In addition to analyzing the image data, I am also learning 2D patterning, micro-fabrication of PDMS masters using lithography, and culturing mammalian cells in vitro. It is all very interesting also a flood of information. For the moment I am learning the basic steps and making sure that I understand the procedure before I attempt to make a pattern by myself.

Over this past week I have looked at about 60 or so time lapses of cells on 250 by 250 micron squares. I have just finished the control, and after analyzing the data in MatLab, the results shows a shape that closely resembles that of a square with rounded edges. The data is all very interesting and it is exciting to analyze the data as no one knows what the results will be or how different drugs that affect endogenous stress will affect the overall patterning of the divisions.

Tissue Morphdynamics Lab: Week 5

So, its Danny, again, and this post will hopefully be a lot shorter than the last one. The fifth week was when the multiple hours in the afternoon spent dissecting chicken embryos would start to pay dividends in research. At the start of the fifth week I started to culture lungs in tissue culture, which was very challenging due to time constraints. In the previous post I mentioned that lungs would only be preserved in PBS solution for up to an hour, and then after an hours they would be considered "dead". Therefore I only had an hour to dissect four or more day 5 chicken embryos, extract the lungs and place them in culture dishes. Despite having had numerous hours of practice, I never felt pressured in dissecting the embryos as I had all afternoon to practice.

Needless to say, my first attempt at dissecting chicken embryos was mediocre at best, as I managed to get three chicken embryonic lungs in a little under an hour. Then Amira, the research specialist, showed me how to culture the lungs in wells. Additionally, all the work was done in a chemical fume hood and all surfaces, bottles of media|serum and gloves were sprayed down with 70% ethanol in order to prevent contamination.

To make the media that the lungs would grow in, we added 10mL of Hyclone DMEM/F-12 (1:1) glutatmine and 500µL of FBS serum into a 15mL sleeve tube. In order to mix the solutions, we simply turned the sleeve upside down multiple times, instead of vortexing it. We then got a 6 well plate and added 2mL of FBS|DMEM solution into a well using a 5mL pipette. Next we placed a single Whatman Nucleopore Track-Etch Membrane into the well with forceps, with the glossy side of the membrane facing down towards the bottom of the well. The Nucleopore Track-Etch membranes ensure no contamination, have a smooth flat surface that allows for high visibility of particles during a microscope, and have high chemical resistance. The membrane essentially sits suspended in the media. Lastly we used a transfer pipette to move the three lungs from the PBS solution on the membrane. The lungs are then grown in a carbon dioxide incubator that cultures the epithelial cells. We take pictures of all three lungs at 0 hrs, 24 hrs, and 48 hrs, which will later be used for analysis.

Day 5 Lungs (0 hrs)

Day 6 Lungs (24 hrs) 

Day 7 Lungs (45 hrs)

The purpose of growing the lungs in tissue culture is to analyze the results using ImageJ, a program that analyzes the area, perimeter, and other useful measurements of images. This data is then compiled onto an excel sheet, which is in turn used to create a graph that can be used to determine the tendencies of lung development. When the lungs are further treated with different concentrations of lungs, these graph also have varied results. The following are images of both the analysis I have done so far as well as the work that the lab has already done and published in an journal. In the excel sheet graph, the x axis represents the area of the epithelium, whereas the y axis represents the number buds located on the epithelium. The results also include all date from 0hrs to 48hrs.

Lung Development Graphs
Source: Jason P. Gleghorn et al (2012)

Lung Development from a few samples.

At the end of the fifth week, the microscopes were finally available and we were able to image the lungs I had stained using a phase contrast microscope. The confocal and phase contrast microscopes at the lab were to be used with extreme caution, as they had broken down multiple times over the past two months, according to Amira as well as other post-docs in the lab. Therefore I watched Amira image the lungs using HC-Image Live, and found it very interesting. Using a transfer pipette, we placed the lungs with PBS solution onto glass cover-slips and then using a beam of light, we found the optimal wavelength (around 460 nano-meters) that produced a good quality image. It was really interesting to see how the microscope was used, as there was a lever on the side of the microscope that changed the frequency of the light. The images are in black and white, but in the live feed from the computer they were a nice turquoise.

Day 5 Embryonic Chicken Lungs

Day 7 Embryonic Chicken Lungs

Weeks 3-4 Mendelsohn Lab

Hi this is Jason from the Mendelsohn lab at the Columbia University Medical Center. Compared to my first two weeks at the lab, I have been given a lot more work to do. For most of my time, I have been able to section many paraffin blocks embedded with either a mouse embryo or a mouse bladder using a rotary microtome. By cutting and collecting the important tissue, microns in thickness, we are able to take an even clearer look at the development of bladder cancer. I also have been able to stain many of my slides using the H&E staining method or by the immunostaining method, applying specific antibodies. Yesterday, I witnessed my first mouse sacking. Katya brought one of the mice to the lab in a small white box. I didn't know what was inside until she told me. I was startled when she opened the box and held the small black rodent in her hand. Then she put one hand on its neck and the other on its tail and pulled, dislocating its head. I was really freaked out but tried not to show it. She laughed at my attempt to conceal my emotions. She cut open its abdomen and took out its bladder. I was amazed. I have dissected a number of animals in the classroom, but I have never witnessed the actual sacrifice of a live animal to be dissected seconds later. It was a whole new experience for me.

I can't believe its been almost a month since I first started and the time spent at the lab has been flying by. Overall, it has been a great experience working in the lab.

(Paraffin block-mouse embryo)

(One section of the block in the water bath) 

(1 1/2 weeks of cutting and staining)

(Microtome and water bath) 

Tissue Morphodynamics Laboratory: Weeks 4

Hi, this is Danny with a couple of updates regarding my progress in Dr. Nelson's Tissue Morphodynamics Lab. My next three posts are going to be a load of information, as I learned a lot of new material and methods over the course of past three weeks.  I understand that these posts are little late, but it could not be helped, as my laptop with all images crashed. I felt that the words I was writing would not do the  techniques or methods any justice so I figured I would wait for my laptop and my images. Just to note, these three posts are actually just one post, but I just prefer to have the weeks organized as there is going to be a lot information being posted.

My lab bench where I stain, aspirate and fix the lungs.

One of the three dissecting microscopes I use.  

The fourth week at my lab was definitely more exciting and productive than previous weeks, as there were no interruptions or unforeseen obstacles. In addition to a new high school student, Chris, from Montgomery High School joining the laboratory, there were also three four new senior thesis students and an undergraduate student. Needless to say the lab got crowded very quickly and I found myself changing desks numerous times. Although the dissecting microscopes were usually occupied during the mornings, I found myself with microdissecting chicken embryos the entire afternoon from mid-day to around five. My first day of microdissecting day six chicken embryos was somewhat embarrassing, as I microdissected ten or so chicken kidneys instead of the lungs. Having been told that the lungs are slightly larger in day six embryos than day five, I found myself with a plate of chicken kidneys. Although ten or so chicken embryo unfortunately were lost, I found that microdissecting chicken lungs was a lot easier than trying to take out the kidneys. Around the middle of the week I already had fifteen or so day 5, day 6, and day 7 chicken embryo lungs waiting to be fixed, stained, and imaged. 

Although I went into a little detail regarding the staining of the epithelium in my previous post, I figured that I would go into a little bit more detail, as I see other EXP students are staining as well.

Throughout the microdissections and the staining procedures, Phosphate Buffered Saline (PBS) is crucial in order to preserve the lungs from an hour up to months. After successfully dissecting at least six chicken embryo lungs I head over to my lab bench with my dish of lungs submerged in PBS and start to fix the lungs. In order to make the formaldehyde solution for fixation, I add 10mL of stock 16% paraformaldehyde (PFA) solution to 30mL of 1x PBS, resulting in a 4% PFA|PBS solution. Then I add the lungs into a well, usually three or four to a well, and then aspirate the excess PBS solution using an aspirator and add the 4% PFA|PBS solution to the culture dish. I then place the dish on a shaker for amount 15 minutes and then aspirate the PFA|PBS solution and washing the excess PFA|PBS solution three times with PBS. Next, I prepare a Phosphate buffered Saline with Triton X solution (PBST) by adding 1.5mL of Triton X into 500mL of PBS, and add a stir bar and leave it on a stir plate for about 30 minutes in order to make a 0.3% PBST solution. Triton X is a very viscous solution and is used in order to make holes in the tissue and mescheyme, which allows for antibodies to react with the epitopes for primary binding in staining. To continue fixing the lungs I add the 0.3% PBST solution to lungs for about 15 mintues on a shaker. Then I add a blocking buffer, which consists of 4mL of 10% goat serum (GS) and 30mL of 0.3 PBST, in order reduce background noise when staining and imaging. I leave the lungs submerged in in the blocking buffer in 4°C for four hours on the shaker, as GS tends to attract bacteria if left at room temperature. After four hours I add the primary antibody, which consists of 1mL of PBST and 5µL of LCAM. After four hours and aspirating the blocking buffer, I add the LCAM primary antibodies overnight at 4°C on the shaker. When I return the next day, I aspirate the LCAM solution and then wash the lungs in PBST for about 4 hours, aspirator and adding sterile PBST every hour. Then I add the secondary antibody, which consists of 488 nano-meter (turquoise color) goat-a-mouse serum. I wrap the plate of wells in tinfoil and leave it overnight. When I return I usually spend up to two days constantly aspirating and adding new PBST to the well of lungs, in order to prepare for imaging. As one can see staining takes about half week at a time, so it is best to have many lungs to stain at once than repeat the process multiple times. 

Although the fourth week mainly consisted of microdissecting, messing up microdissections, and then staining the lungs in order to prepare for imaging the following week. I also learned a lot more about the work others were doing in lab through the group meetings, which consisted of Oscillatory Rotation and Cell Motility and the Mechanics in Meschyme-free branching. What was interesting in the latter topic was that the post-doc cut the distal tips of chicken embryos, which reportedly are places of high stress and proliferation, and embed them in matrigel which was submerged in fibroblast growth factor (FGF), a family of growth factors that are key in embryo develop. Using time lapse imaging on the confocal microscope, he was able to create a animation of the terminal end buds actually growing out of these square matrigels in culture. Although similar works has been done, which I will provide an image in a second, I don't believe anyone has necessary done the work recently using modern technology. 

An  general idea of the project discussed during the group meeting.
Source: Fibroblast Growth Factor 10 (FGF10) and branching morphogenesis in the embryonic mouse lung

Lastly, in between staining and microdissecting chicken lungs, I was busy learning other useful techinques that may or may be useful in the near future. I learned the process behind making complementary DNA (cDNA) and quantifiable polymerase chain reactions (qPCR), which is similar to the polymerase chain reactions (PCR) we did in AP biology. In short, the entire cDNA process consisted of following a instruction booklet sent from Thermoscientific Fisher and numerous attempts to find a optimal amounts of solutions to add. qPCR was similar to PCR except that the procedure would allow for one to quantify the relative amounts of increase in DNA overtime relative to other samples using spectrometer, whereas in PCR one could only analyze the amounts of DNA after running the DNA through gels and imaging them in ultraviolet light. 

The fourth week was a lot of new and exciting research, as well as new and exciting methods and techniques. I learned a lot during the fourth week and got to know a lot of the people in lab, besides Amira and Ben. 

The Wait Finally Ends

Hey guys, it's Katie from the Children's Hospital of Philadelphia again.

So last week I received some news that I thought would be good, our probe was ordered and should arrive any day!  On the day Dr. Peretz and Dr. Crider came to visit my PI told me that even if it didn't arrive I would be able to start my project.  Unfortunately, that didn't end up happening and I was told to wait for delivery.  Almost a week later, we found out that the probe would not be arriving at all and I was given the go ahead to start my project.  With only a week to do a month's work of work I haven't had much down time.  Our project includes extracting, plating, and analyzing over 200 patient samples for HSV 1 and HSV 2.  The samples are run through a real time PCR thermocycler for amplification and the results are presented on a computer system.  I can't show you on the blog because I don't have access to my pictures yet but I will come poster time.  Anyway, the plate we are using has 96 wells all of which are filled with 50 mL of liquid.  Picture a tear drop in a miniature thimble and you kind of get the idea.  Our trial includes 4 different assays, two of which are type common (don't differentiate between HSV1 and HSV2) and the others are type specific (they tell you which one it is).  Each assay has a different target sequence so we are comparing the sensitivity and specificity of each.

My days now consist of arriving at the lab, retrieving my samples from the refrigerator, and moving to the amplification area.  Plating one run takes about an hour because you have to wait for your samples and master mixes to thaw, spin down your master mix, and then proceed to the actual aliquoting  of sample and master mix.  I generally get 2-4 plates done a day and I am down to my final run today.  By the end of today I should have my data ready to be analyzed and I will begin the long process of condensing a binder full of material into four spreadsheets.  Sounds fun right?

Overall I am enjoying my time in the lab.  Although I had more downtime and waiting time than anyone originally expected my project is coming along and things are looking pretty good. 

More Dimers - Week 3

Hi I'm Harry and I'm working in the Ballatore lab at UPenn, I'm joining in the lab's project to create a set of carboxylic acid isosteres.

After finding the dimerization issue, we tried two different things: 1. Do the same alkylation with the left-over CPD but with a 1:1:1 CPD base and benzyl bromide ratio, we did this because the dimerization might have been caused by excess CPD and not enough base and benzyl bromide to react with. 2. Do the same alkylation but this time with a methylated version of CPD. My PI predicts that the methyl group should prevent the formation of CPD dimers. The LC-MS of the product of both methods showed significantly greater yield of the product we want but unfortunately we still see the dimers. We deprotected the products and using high-performance liquid chromatography (HPLC), which accurately separates our desired product and the impurities, we collected our products.

To deal with the persisting dimerization issue, my PI suggested we change the base from KHMDS to LDA, a stronger base because he had done an akylation with LDA without encountering any dimers. To do this I had to make more protected CPD. Meanwhile we set up a separate reaction with a completely different starting material. However, the reaction did not produce any of our desired product and was considered a failed reaction.

This is the general outline of the week's lab work but also a lot of time was used in regular column chromatography, extraction, and the most annoying part: waiting for the extracted product to dry. I'm starting to repeat the same procedures for each reaction: reaction setup, monitoring using TLC and LC-MS, then quenching/extracting with water and ethyl acetate, chromatography (HPLC or regular chromatography), and finally an NMR analysis. It's great that I'm finally starting to do these steps by myself without my PI or post-doc.

The End is Quickly Approaching- Week 5

Just a reminder, my name is Colton and I’m visiting the Buccella Laboratory at New York University. The lab is part of the chemistry department. More information about the lab can be found @ http://www.nyu.edu/fas/dept/chemistry/buccellagroup/ .

As I mentioned in last week’s post, we were still testing the magnesium indicator in the presence of ATP. This week we attempted to determine whether the compound was photo bleaching and/or taking a while to equilibrate. Monday we started two experiments to attempt to answer this question.

To test the amount of time to equilibrate we took fluorescent spectra every 10 minutes, allowing the solution to stir and settle between spectra. This allowed for any complexes to form between the sensor, magnesium, and ATP. The fluorescence decreased as the solution equilibrated, which we thought may occur since we believe that ATP was quenching the fluorescence. However, we let another vial of the same solution sit all day and equilibrate by it, and when testing it at the end of the day, it had the same fluorescence level of the first spectra taken. This indicated to us that the solution freshly prepared does not take long to equilibrate, since it has the same fluorescence as a solution that was able to equilibrate all day. Thus, the steady decrease we saw every time would mean that the compound was photo bleaching.

We decided to do a time spectra for to test for photo bleaching. Every 2 minutes the fluorimeter automatically took spectra of the compound in buffer. This frequent exposure to light will break the compound down if it photo bleaches and show a slight decrease in fluorescence each time spectra is taken. Our assumption that the solution was photo bleaching seemed to be true, until approximately the 19^th scan where fluorescence began to increase with each additional spectrum taken! We were baffled by these results.

More or less, we were still left stranded trying to come to some sort of conclusion as to what the system was doing. Our only option left was to do an NMR titration to know what complexes were forming in the solution and hopefully answer why the spectra was going down then up.

To make sure we didn’t get incorrect data we repeated the experiments. We will not be able to start the NMR titrations until next week.

We went in this week with a lot of questions, and even though we didn’t answer any, our data will have meaning. After we figure out what complexes are forming we will hopefully be able to understand this week’s data and previous week’s data. I’m not looking forward to the start of next week, since it will be my last 5 days in the lab. I have learned so much about fluorescence and science in general. It won’t just be the end of a summer internship but also the end of many days spent with great people and many laughs all within the realms of chemistry. 

Week 4 at the Donohue Lab

Hi this is Meg again, writing about my fourth week working in the Donohue lab at Duke University, where we are researching the different effects of water potentials on seed germination. As I described in my last post, on Monday we finished rinsing seeds from Lien’s project. Then for the rest of Monday we had to census seeds from both from my experiment and Lien’s.

Tuesday morning, we had to head out to the field site in order to weed so that other plants wouldn’t affect the growth of the experimental plants. However, the field is ridiculously overgrown, with weeds taller than I am. As a result, the weeding was pretty intensive and it took us 2 hours to clear out a tiny area. Also, because North Carolina can get extremely hot and humid by around 9 or 10 in the morning, we went out into the field at 6 am and worked until 8:30. Once we finally stopped working, I went back to my apartment to change out of my muddy and dirty clothes and then headed over to the lab. For the rest of the day I censused our experiments. After lunch though, Lien and I met with the postdoc, Tarek, to discuss the analysis of our data. Basically this ended being a 2 our graduate level statistics class, but I still successfully followed along, and was pretty interested by the end.

Inside of chamber with plates

Chamber where plates are kept

On Wednesday, luckily we didn’t have to go to the field again. However we still had to census all day. In addition, we had another 2 hour meeting about analyzing data.

Then, Thursday morning, it wasn’t required, but Lien and I volunteered to help in the field again, meaning another really long day. When we returned to the lab, we censused some more. Then just like the other days this week, we had another meeting. This one was the longest yet and it went over 2 hours and 30 minutes. By the end of the meeting, Lien and I were visibly exhausted, so Bri let us leave an hour early.

Microscope used for censusing

Sheet used to census

On Friday and throughout the weekend, we just censused during the day since we had finished data analysis and field work for the week.

Once again, this week has been exhausting. However, all this work has been pretty enjoyable and has helped me to stay occupied in the lab. As I said last week, I prefer these busy weeks to the fairly slow weeks that I had when I first began working in the lab.

Blueberry picking, lunches and final presentation - Week 5 at Duke

This is Jocelyn and I'm going to be talking about my final week of EXP. Just a
reminder, I've been working on olfaction in mice at Duke University.

My last week at Duke was a quite surreal. Neha's flight back to Durham from
California was delayed until Monday night, so on Monday I worked on my
presentation. Ting presented her paper for journal club on Oxygen Sensitivity of
mitochondrial function in rat arterial chemoreceptors. It was pretty interesting.
Tuesday and Wednesday afternoon, Neha helped me with picking the appropriate
images to show for the presentation and analysis and graphing of my data.

On Wednesday morning, the whole lab went blueberry picking! It was awesome.
People who didn't have cars got lifts from people who did at the lab at 9, and then
we all met at the blueberry farm at 9:30. Yue drove me and Sahana (another high
school student), and we got lost for at least half an hour because there was no
sign for the blueberry farm! Anyway, we got there and picked blueberries. It was
really cheap too. We stayed at the farm for a bit, and then eight of us went to the
diner for lunch. This was when I really saw how much support the lab members
had for each other and how much they appreciated some "down" time.

On Thursday, it was my "big day". I presented my data and findings to the whole
lab with Neha by my side. Hiro pointed out some mistakes I had made in defining
and the results that I presented. Neha and I talked through them and she's going
to count the positive cells again. The process is really subjective because some
may count one cell has positive but others may see it as just the background, so
she's going to figure that out for me. I brought some dunkin donuts for everyone,
and Jessica (the lab technician) brought some brownies, cookies, green tea and
chips. Hiro, Natalie, Ming-Shan had some questions and I answered them with
some hesitation but with the help from Neha, it was all good. It was sort of
intimidating at first because I was presenting to very intellectual people but they
were all very supportive. Hiro said that he wasn't going to be in the lab on Friday
so he was going to say goodbye. I told him it was truly a great experience
working in his lab and we're going to keep in touch!

On Friday, Neha and I talked about the results and the countings. Then she took
me out for lunch at Sushi Love. Ming-Shan heard about it and asked if she could
come. Just five minutes later, the whole lab was walking to the restaurant. It was
really cool. Then at five thirty, I said goodbye to everyone. It was really an
amazing experience working at Duke. To be honest, I miss it!

Here are some photos from the past five weeks:

This is one of our gels for genotyping. 

This was my Duke ID card. It gave me access to the lab.

This was my lab bench area. 

This was the microscope we used for looking at the slides in different filters. 

I used this cryotome to section paraffin blocks of tissue.