Have a great end of week and beginning of next week. I'll be in Alaska - www.mbari.org/EARTH and back in the office either Weds afternoon or Thursday am. Don't forget the Geology field trip on Monday - 10 am!! be there, it's a great trip.. you'll have a blast. you'll be back by 3 at the latest.
Also, forgot to say anything but next weekend is the Scottish Games and Celtic Festival
http://www.montereyscotgames.com/

have fun

I've only been in lab a little over an hour and it's already a great day. I had an incredibly easy calibration of the Raman this morning, which entails focusing the laser on an almost invisible silicon wafer, and am now off to a running start with some pretty cool spectra. Yesterday I made up several ternary gas mixtures of CH4, CO2, and N2 to test the interactions of these three gases together under varying pressures, which I am now analyzing with the Raman laser. In the grand scheme of things, my project is a reimaging of the wild (and theoretically wonderful) hypothesis that CO2 can be substituted for CH4 in methane hydrates on the sea floor, allowing for the mining of these inner hydrocarbons and the deposition/storage of CO2 - a win-win situation that is unfortunately, more chemically and physically complicated than the binary replacement model currently proposed. There are several intermediate steps that need to occur (or so I'm claiming) before this substitution is possible, and one of them entails an impure mixture of CO2 and N2 gas being forced into the hydrate framework, rather than pure CO2. I have produced evidence that chemical changes occur when all three gases are present simultaneously that fundamentally change the substitution mechanism in regards to pressure,temperature and phase. I have yet to elucidate exactly what those reactions are, and hopefully today/tomorrow that will be sorted out and my proposed revision will be solid. That's the hope anyway, and time is running out! I can't believe next week is my last week at MBARI before flying out to Oregon.

I have spent the bulk of this week prepping for our time on the Flyer, which has been an exciting look into what it takes for a research cruise to be successful (a lot of time, organization, and mind-reading by Peter Walz and Ed Peltzer as Peter Brewer is on vacation in Scotland)! I tried to sneak my way into the driver's seat of a forklift, but so far to no avail since I don't have a special license. I'll keep at it. Next week, I will be finishing up these tertiary data/writing my report/finishing my presentation/packing everything for the Flyer/packing for Philadelphia/entertaining my mom/spending as much time as possible in the Pacific before I leave. Can't wait! 

This week has been rather successful, even if it feels like I haven't gotten that much done. The Browse Taxonomical Tree page is (finally!) working again, and I have successfully implemented an algorithm for generating representative images for higher level concepts by selecting random images from the children. It is extremely slow because it pulls a lot of data from the database, but Brian told me this morning that he has a way of writing what I want to do but much faster. He's taken off today, so I'll have to wait til next week for him to do that, but that's alright. I've also made a lot more layout changes based on suggestions from the video lab and still have a few more to make before we meet again. I've also been taking more screenshots for my presentation.

Next week I will be working on either implementing autocomplete on all search boxes or making the tree collapsible/expandable in a more tradition way (think expanding folders in sidebars on most standard PCs with the little +'s and -'s on the left of the folder name). Or both, if I have time!  I'd also like to get an old version of the DSG running on my computer so I can take some screenshots of what it looked like before I started. I am also going to work on adding a link to a histogram showing the organism's depth distribution in the comparison table and adding other properties the video lab wanted as well...it is going to be a very crowded table. If I have time, I'd like to make the table customizable by column as well as by row so the user can only look at the properties that they're actually interested instead of always having a cramped table with columns the user doesn't need. All of these are time-based, of course, so I will do however many of these ideas as I can before August 19th. And of course I'll start working on my paper and presentation.

This week I started some preliminary data analyses. I got a refresher course in the computer program Arlequin, and ran AMOVAs and a few other statistical analyses to get an idea of the genetic diversity amongst the individuals I have sequences for, and to see how that diversity relates to the biogeography of the tubeworm. I also made some preliminary parsimony networks, which gives pretty much the same information, but in a pretty (and sometimes not very pretty) picture. It looks like there might be more going on with Tevnia than a related tubeworm recently studied, so hopefully I have some interesting things to talk about. These analyses also revealed the latitudes for which I need a larger sample size, so I am now going back and redoing COI and cytB PCRs and sequencing reactions for those that did not work the first time. I will get those back next week, edit them, and then I should have a complete data set for those two genes (because they will work!!). I am also still really hoping to have data for a nuclear gene by the end. I have two primer sets that have produced results for a small sample size, and I should have the sequences by next week when I will really know. There are doubts that one of these genes will be polymorphic, but one looks promising, if the sequences come back clean enough. I will find out next week.

Break down of my week:

Monday- don't remember Monday.......... moving on

Tuesday- pretty much consisted of asking "God, when you planted the idea of statistics into the mind of the guy that invented it, couldn't you have added a little foot note reminding him not to make it so confusing?" That and bugging Grant for help with graphs- help with I am quite grateful for.

Wednesday- banging my head against a wall while making more graphs- then making even more graphs with Paul- Halleluiah! a breakthrough!- Then making tables for the statistics I translated the day before and emailing it all to Linda.

Thursday- Getting answers from Linda, doing more work, getting muddled, getting the muddle fixed, getting muddled again, emailing Linda, and hoping my fix to this last muddle was right. Now I am coming up with what I think the graphs I made are telling me and getting further into the paper.

Next week- paper and presentation Continued. 

So far this week I have accomplished more than I originally set out to achieve. I was attempting to integrate other orders of ctenophores into the key (to supplement the cydippids already present) and I have added almost all known genera of ctenophores now. However, due to the extent of functional differentiation observed between some of the cydippids and other groups like cestids, there have been a number of changes to the questions I will ask in my key. these changes have knock on effects that mean I will have to rescore the entire key, once I am sure on how I wish it to be structured.

I have added screenshots (and very simple vector art) to the features list to show what the specialist morphological terms (substomadael meridional canal, infundibulum etc) mean. This should help make the key more intuitive.

Now that almost all the genera are added and preliminarily scored, the structure is more less set so I have begun the major rescoring. I should have that finished by either today or Monday, leaving the key in a pretty complete state.

The upcoming challenges include successfully uploading the key to the internet (which has proven complex so far), finding a way to export the key into taxonomic software capable of constructing phylogenies (such as Mesquite) and ensuring that the key is easy to use and accurate regardless of the level of knowledge of the user.

This week the batch classification results have been processed and detected ALOT of events. Knowing full well that the majority of it will be noise, it is still interesting nonetheless how many frames AVEDac regards as 'interesting'.

I began the 2nd part of my analysis which I call Man vs. Machine where I would annotate a small randomized sample (5%) of the entire data set and compare with the classifier. I had planned to carry out the more viable 10% sample test, but time is of the essence and some of the clips had a lot more events than others! I created a randomized number generator out of all the clips using functions in Excel. The concept of Man vs. Machine will hopefully give us an insight as to how well AVEDac (and myself) can perform and classify what we 'regard' as correct. This is one of the most fascinating parts of my project because it is involving a lot of deep thought about how we and machines are actually quite similar in terms of the way we learn. The results will hopefully help clarify this idea!

I finished my annotations today, extracted all the confusion matrices and table of events and now currently pondering a way to combine all of this together. Today has been running on a steep learning curve in terms of getting Excel to do what I want it to do. There are a lot of data points, but aiming to try and get it complete by the end of today. It seems I am doing it an old school way according to my mentor, as she is working out how to use a magical matlab code which SHOULD do it all for me!Its times like these I wish I had a degree or some sort of intuitive experience in programming.watch this space...

Next week will be spent mulling over the results and interpreting what it all means!I will hopefully have finished my methods section by the end of the weekend, leaving only the crucial part of my report to write up.

This is the last fourth working week for summer interns. This is really a good opportunity for me to do research and focus on the project. Appreciate this chance and appreciate all the help from MBARI
This week, I am focusing on the hardware study, wire the McStar instruments and do some laboratory test. Collect some acoustic data. Based on Linux system and window system, I try to plot the acoustic signature spectrum and look in detail about the laboratory database. Some unknown sound was collected by hydrophone, 60Hz. We guess it is from the electronic noise in the lab power supply. So we set up two batteries case for power supper taken place of prior one. Also, with the help of electric staff, I wire new cables for tests. When I have some data, I use Matlab to analyze the spectrum, and find some issues about the setup, the input signals are not stable and not reasonable as expected. Then, I tried several ways to rewire and do tests again. Find some problems: batteries issues, cables issues and computer collection problems. After solving these problems, the test is ready to continue.
Thursday, I spent the whole day to read some documents, tried to understand previous research and programming codes, and schedule the goals for next weeks. At the same time, I do some research on acoustic classification and detection, tried to get more information about acoustics, which will be more useful and helpful for my further understanding.

Next, we propose to do tests in the tank and get the real test data. It is the last third week, I must catch up with the schedules and try to get some work done on time.
To do:
1. RE-do the laboratory tests to make sure everything goes well.
2. Do the tank test and collect the real data.
3. Make MATLAB code running and easy to use the code to analyze the acoustic signature.
4. Find the 60 Hz acoustics and do some research on that, try to find out what it is and where it is from.
5. Think about how to filter this acoustic noise if it is from power plants or some else. If it is from underwater ocean floor, that will be much interesting for the further research.
6. Plan to do the flied test in the following week.

At the end of last week Lonny showed me how the MatLab code works for creating computer mosaics so I've been playing around with that a bit using some images of the USS Macon, an airship that went down off the coast in 1935. Making these mosaics takes quite a lot of computer time but can be left running in the background so long and will just throw up an alert if it needs human input.

So while I left this running I exported all of my data from VARS, cleaned it up, and turned it into presence absence data for all of the taxonomic groups I'm looking at. The computer was still running the mosaic code by the time I had done this so I also taught myself how to make basic 3D images using ArcScene which is really nice in terms of looking at the 3D bathymetry models.

For future data analysis it's looking like we'll be running a PERMANOVA rather than numerous ANOSIMs due to the heterogeneity of submarine canyons and the large number of variable to account for. I'm still reading around all of the info on PERMANOVA to see whether it is what we want or not. It's also looking like we'll be taking random sub-samples of each transect for analysis, although technically pseudoreplication this will hopefully increase our options for analysis.

This week was very exciting! I received the rest of my Atolla slides back from CHOMP and was able to compare the trailing tentacle with the marginal tentacle. There are some very interesting differences and similarities, both of which will become more meaningful once I can decipher, on a deeper taxonomic level, what I am in fact looking at. I spent a large chunk of this week doing research in the library, joining the cnidarian listserver, and emailing various jelly experts and enthusiasts in order to get some sage advice on identification of tentaclar ultrastructure. Every day I am getting closer to enlightenment on this situation, but even better would be if I could get some samples through an SEM or possibly even a TEM, to get a better look at these cells and structures. Next week I will be figuring out how to prepare a soft bodied organism to run through an SEM, which will hopefully produce micrographs that can assist me in the identification of nematocysts and other external structures on the Atolla tentacles. Wish me luck!

I have to do lot of work in this week to complete my lab work. so I started with another set of cleaned specimens. I managed to mount all of these on epoxy.Then on wednessday I polished them using polishing equipment at moss landing marine laboratories. by thursday I started to observing them, but couldn't complete all. the most difficult task is to distinguish different bands. This week I'm Planning to finish alll of my lab work. I have already started my paper too.

This week things could have gone more smoothly.  On Monday, having completed all of the major wiring, I connected tried to power up the electronics, however the equipment was unresponsive. After some trouble shooting we figured out that the OASIS 4 board may have been shocked and that the Chamber Controller board was also not working properly. On Tuesday and Wednesday I began to write and organize my paper while I waited for a new Oasis 4 and Chamber Controller. On Wednesday I also powered up the Wave Gliders communications system and successfully made contact with the Wave Glider via an xBee radio communication.  After establishing communications I received two error messages.  After contacting Liquid Robotics and trouble shooting the error I was able to resolve the issue. On Thursday the Chamber Controller and OASIS 4 boards were ready. This time the electronics turned on, however I received several error messages. Resolving these errors will be my primary goal for this week, however it two of the valves did work, so at least some of the parts are working. On Thursday I also made contact with Liquid Robotics in order to set up on communications via the internet.

Lots to do next week but hopefully resolve all of the errors and successfully have our electronics communicate with the Wave Glider communications. Some additional coding will also have to occur because we are adding a CO2 standard which has yet to be used with our OASIS controller. 

This week I've been working on displaying pH calculated from seasurface temperature, and modelling how it will change with different climate situatuions. This has involved lots of reading and thinking, then reading some more, then asking questions.

My next task will lbe to get get my new fancy color map working, then applying the relationships that I found.

Last week I collected the data necessary to write up the results section of my paper and finished a draft of most of the rest of it. I also did some comparisons with the LCS data, but haven't figured out a way to formalize that yet.

 This week will be finishing the first draft of the paper so that we can finish editing it for submission before I leave. I also have a couple of other measures that I am going to apply my algorithm to to try and solve some of the edge cases that it doesn't work so well on yet.