This is where things get funny. Outperforming my home page by nearly double was indexOld.php (now indexOld22.php) – a simple webpage I tossed of for about a year before I put my big blog back online! Why were people still going to this page? Further investigation (from the referring sites section of my stats page) revealed a lot of hits from Google image-searches. I started looking at the actual requests and realized that many of these hits were people searching for the term Dwarf Gouramis “a type of freshwater aquarium fish) which was mentioned on that old webpage. The ironic part about it is what happens when you google image search for dwarf gouramis there is a picture of an extremely rare zebra pleco which is actually a link to my website! However the link APPEARS to be to wallpaperfishtalk.com because on my page I just linked to their image.

My conclusion: People are Google image-searching for ‘dwarf gouramis’, and an amazing picture of a zebra pleco is coming up which links to my site (due to the fact that months ago I talked about dwarf gouramis but posted a photo of a zebra pleco) and people (in their awe at this amazing fish) are clicking on it. So what did I do? I pulled a bait-and-switch! You bet I did. Now when you go to indexOld2.php it just forwards you to my current website – mua ha ha ha ha

PS: I’m appending to this entry at 2:17pm to note that I made a wonderful breakthrough in the lab today. Due to intellectual property protection blah blah and the fact that I don’t want anyone else to beat me to my research goal I will not describe what this is, I’ll just say that it took months of preparation and today – presto! It worked beautifully =oD

*With all of the uncertainties already afflicting my thesis process, I didn’t want animal confusion to be a problem.* From what I heard, about a dozen diabetic animals (in the age range I wanted most for my experiment) were recently killed. I felt that this process needed to be organized before I could properly come up with a plan, and of course, Python would be involved. I spent the morning downstairs manually inventorying mice. I decided that a location-based system would be the most useful. Each cage got its own line on a sheet of notebook paper (actually more like 8 sheets), and on each line I noted the cage location (in story/row/column coordinates), and information about the 1 or 2 groups of animals housed in the cage (count, sex, and date of birth for each group). After a couple hours sweating beneath my disposable hairnet and lab coat, I completed the downstairs inventory. To the side is a picture I took of myself half way through the monotonous process. (I think blood was about to shoot out of my eyes) Later I ate lunch, then did the upstairs. After I finished, I went home and manually entered this data into the computer (open office spreadsheet -> XLS), so it looked like this:

*Then I saved this data as a CSV* (comma separated values) file (using either open office or microsoft excel). Note that this is done in such a way that it’s very easy to make modifications to the XLS file to reflect new cages, changing cage locations, or animal death. Anyway, once it was in a CSV format, I wrote a Python script to convert the CSV format to a custom easy-to-analyze format which basically boils down to one line per mouse (not one line per cage). This is the Python script:

f=open("data.csv")
raw=f.readlines()
f.close()
animals=[] # [cage,sex,type,dob,age]
cages=[]
def addAnimals(animal,cage):
        if len(animal[0])==0: return False
        if cage not in cages: cages.append(cage)
        count=int(animal[0])
        if animal[1] == 'm': sex="male"
        else: sex="female"
        strain=animal[2]
        dob=animal[3]
        age=int(round(float(animal[4])))
        animal = [cage,sex,strain,dob,age]
        for i in range(count):
                animals.append(animal)
for i in range(len(raw)):
        line = raw[i].replace('"','').replace('n',"")
        if "-" in line:
                line = line.split(",")
                addAnimals(line[3:8],line[14])
                addAnimals(line[8:13],line[14])
f=open("data.txt",'w')
f.write('['+str(animals)+"],["+str(cages)+']')
f.close()

*Running this script produced a file with content like this:*

[[['d-a1', 'male', 'fvb', '09/26/08', 3], ['d-a2', 'male', 'fvb', '08/12/08', 4], ['d-a2', 'male', 'fvb', '08/12/08', 4], ['d-a3', 'male', 'fvb', '12/24/07', 12], ['d-a3', 'male', 'fvb', '12/24/07', 12], ['d-a4', 'male', 'fvb', '11/16/08', 1], ['d-a4', 'male', 'ove26', '11/16/08', 1], ['d-a4', 'male', 'ove26', '11/16/08', 1], ['d-a4', 'male', 'ove26', '11/16/08', 1], ['d-a5', 'male', 'fvb', '10/26/08', 2], ['d-a5', 'male', 'fvb', '10/26/08', 2],...

*From there, I wrote an analysis python script.* Should I post it? [thinks about it] Sure why not. A warning though, the code is pretty rough. It works though =o)

import pylab

f=open('data.txt')
animals,cages=eval(f.read())
animals,cages=animals[0],cages[0]
f.close()
print "Processing data for %d animals in %d cages..." % (len(animals),len(cages))

##LIMITS######
sex="male" #'male' or 'female'
strain="ove26"
minage=3
maxage=9
##############

selected=[]

def passIt(animal): #['u-l7', 'male', 'ove26', '10/28/08', 2]
        global selected
        if not animal: #dscription
                title="Displaing "
                if sex: title = title+sex+" "
                if strain: title = title+strain+" "
                else: title = title+"all "
                title = title + "mice from "
                if minage: title = title+str(minage)+" months to "
                else: title = title+"birth to "
                if maxage: title = title+str(maxage)+" months."
                else: title = title+"death."
                return title
        if sex:
                if not animal[1]==sex: return False
        if strain:
                if not animal[2]==strain: return False
        if minage:
                if animal[4]maxage: return False
        selected.append(animal)
        return True

def histIt(strain="all",col='k',label=True,limitTest=False,lw=None):
        ages,xs,a=[],[],{}
        for animal in animals:
                if limitTest:
                        col='b'
                        if passIt(animal):
                                ages.append(animal[-1])
                                if not animal[-1] in a: a[animal[-1]]=0
                                a[animal[-1]]=a[animal[-1]]+1
                elif strain in animal or strain == "all":
                        ages.append(animal[-1])
                        if not animal[-1] in a: a[animal[-1]]=0
                        a[animal[-1]]=a[animal[-1]]+1
        for x in range(max(ages)+1):
                xs.append(x-.45)
                if not x in a: a[x]=0
        if not limitTest:
                pylab.title("Ages of %d (%s) Mice"%(len(ages),strain))
                pylab.xticks(range(1,max(ages)+1))
        if not label: col,lw='0.9',0
        else: pylab.title(passIt(None))
        pylab.bar(xs,a.values(),color=col,lw=lw)
        for x in range(max(ages)+1):
                if label:
                        if a[x]>0: pylab.text(x,a[x]+1,a[x],ha='center')

def showSelected():
        cages={}
        ids=[]
        m,f=0,0
        for animal in selected:
                if animal[1]=="female":f=f+1
                if animal[1]=="male":m=m+1
                if animal[0] not in cages:
                        cages[animal[0]]=0
                        ids.append(animal[0])
                cages[animal[0]]=cages[animal[0]]+1
        ids.sort()
        x=0
        disp="Total of %d mice (%dm/%df)nn"%(m+f,m,f)

        disp=disp+"Animal Cage Locations:n"
        for cage in ids:
                disp = disp+"%s(%d), "%(cage,cages[cage])
                x=x+1
                if x>7:
                        x=0
                        disp = disp + "n"
        return disp

def getLimits():
        global sex, strain, minage, maxage
        answer=raw_input("Sex [m,f]:")
        if answer=="": sex = None
        if answer=="m": sex = "male"
        if answer=="f": sex = "female"
        answer=raw_input("Strain [ove26,fvb,gfp]:")
        if answer=="": strain = None
        else: strain=answer
        answer=raw_input("Minimum Age [3]:")
        if answer=="": minage = None
        else: minage=int(answer)
        answer=raw_input("Maximum Age [9]:")
        if answer=="": maxage = None
        else: maxage=int(answer)
        print "nn"
        print passIt(None)
        raw_input("npress ENTER to start")
        return

getLimits()
fig = pylab.figure(figsize=(12,8))
histIt(label=False)
histIt(limitTest=True)
masterAxis = [0,pylab.axis()[1]+1,0,int(pylab.axis()[3]*1.15)]
pylab.axis(masterAxis)
pylab.figtext(.4,.85,showSelected(),va='top')
pylab.show()
#fname=raw_input("Enter a name for this image to save it or press ENTER to quit:")
#if len(fname)>1:
#       pylab.savefig(fname+".png")
#       print "nsaved as [%s]"%(fname+".png")
#       raw_input("npress ENTER to exit...")

*Running this code* asks some questions about what type of information I should display. if I have it display all male OVE26 animals (for example) the output looks like this (thanks to matplotlib):

*Here the gray bars are the total number of all animals, and the blue bars are the animals I searched for* (in this case, all male ove26 animals). Comparing FVB and OVE26 charts, I estimated that we had enough male 4-5 month old OVE26 and FVB animals to make the experiment work (about 12 of each group). Searching for male ove26 mice at least 4 months and no older than 5 months old produces this chart:

See how it lists the location of each of the cages and the number of animals I want from each *cage?* For example, “d-I2(1)” means that there is a cage downstairs, along row I, in column 2, which contains 1 male OVE26 mouse 4 or 5 months old. Awesome list generation! Thanks Python =oD

*But wait, couldn’t I have just gone down and looked for 4-5 month old mice?* Yes and no. Yes, it would be easy to find (and mark) these mice, but no because I would not have known to look for them. The goal of yesterday’s little python project was to be able to see at one time everything we have, so I could best determine the criteria of the animals I wanted. Before I made this program, I was planning on comparing mice between 5 and 7 months – something that appears would have been impossible based on the animals we currently have. I can also tell what other experiments I will be able to do in a few months, when some of the mid-age FVB mice will grow older. Additionally, this program is versatile and can be used again and again, for many different projects, with no modification to the code required. Yay Python!

Although I intended to begin dental school in the fall of 2007, the admissions departments had other plans for my future. Taking my initial rejection in good stride, but still strongly desiring to become a dentist, I decided to expand my knowledge of biology and develop whatever other skills I could by pursuing a graduate degree with the intention of re-applying to dental school. … It has been my dream for many years to become a dentist, and following acceptance into dental school I will work hard to become a prominent figure in the community and a great example for all those who have similar dreams. My undergraduate and graduate school experiences have both equipped and energized me to pursue a career in dentistry, and the skills I’ve acquired along the way have prepared me well to pursue my dream of becoming a dentist.

I know, right? Jeez. Moving on, I wanted to note something about my research. Things are progressing nicely, and I’m about to have my boss buy ~600 bucks worth of antibodies (a total volume of ~1100 microliters, or about 22 drops of a clear liquid). It’s crazy how expensive these things are. Yeah, I understand the concept behind the development of polyclonal antibody solutions – but I’ve never seen a detailed analysis of the costs involved along the way. When I spend $23 per drop of some chemical, where exactly does the money go? Anyway, I had a revelation. One of the most complicated aspects of my project is that the signal I’m trying to observe through the microscope is from a fluorophore that emits light around a λ of 498nm to 529nm, and that thick myocardial tissue is autofluorescent in this region. Yes, I might see a tad of labeled fibers, but it’s amidst a sea of background fluorescence! Today (since I needed to order some new secondary antibodies anyway) I decided to investigate exactly which regions of the spectrum were most affected by myocardial autofluorescence. I blasted some thick atrial tissue with the 405 UV diode and took a series of images with a 5nm-wide recording wavelength window shifted by 3nm each, then took the average intensity of each frame of the stack. The result (when graphed) was a pretty spectral representation of myocardial autofluorescence, which was incredibly string in the blue and green bands that I had been using all along. I’m going to order some far-red secondary antibodies, hoping that it’ll help.

After many more hours of writing code in python, I finally have some presentable data. Due to intellectual property reasons, I’m not going to include details about the method, units, or even samples I used. It’ll suffice to simply say that my method seems to be working well, and that the bottom line appears well outside of the upper band (and its respective standard deviation). ALL processing was completed ENTIRELY within Python. I used the python imaging library (PIL) to load data form the TIFF file into a HUGE array, and NumPY to assist with manging the array. Data was then graphed with MatPlotLib and placed directly into a PNG. Here’s an example!

Okay, it’s getting late and I have a project I have to prepare for tomorrow. I know my blog entries have become boring, dry, and overly scientific since I started writing again, but I have to admit that this is just the kind of person I’ve become. I don’t really write about relationships anymore because I’m kinda “set” in that department (with the whole marriage thing and all), and I don’t think it would be a very good idea to go about blabbing all of our personal life together anyway. Beside, these writings are an expression of my thoughts and (supposedly) not intended to be anything more. If my thoughts are boring, so are these writings. [sigh]