DNA on a Rainy Day (continued...)

On a previous post I attempted to extract some DNA from peas and the experiment raised some questions which I'm trying to address below.

I decided to test some of these conclusions/hypotheses by means of a small factorial experimental design using split v. frozen, filtered v. unfiltered pineapple juice and 1 ml v. 3 ml of pineapple juice as factors. Using a saturated randomised experiment matrix, I can evaluate the effects of these three factors by means of just four runs.

Not being able to find split peas, I settled for dried marrowfat peas, a British favourite used in preparing mushy peas (a common accompaniment to fish 'n chips). I used 100 g of these but due to their extreme water absorption (I soaked them overnight, as I did the frozen peas), I had to add another 50 ml of water to the standard 200 ml, in order to be able to blitz the peas into "pea cell soup". So the comparison below is really between 100 g of frozen peas/200 ml of water on the one hand, and 100 g of dried marrowfat peas/250 ml of water on the other hand. All other factors, such as the amount of salt, the amount of washing-up liquid and working time, were kept as constantly as possible. The "pea cell soup" obtained from the marrowfat peas was much thicker and a lighter shade of green with a bit of froth (foam) on top (see also below).

The pineapple juice was obtained by blitzing half of a baby pineapple, passing the fruit paste through a tea trainer and filtering half of that using a funnel and some kitchen towel (tissue). The filtered pineapple juice was slightly opaque and thin, the strained pineapple juice was thick and completely non-transparent.

The experiments were run almost simultaneously, using about four ml of the treated pea soup (recipe)

Here's the experiment matrix and the response variables obtained:

Run>>>>> Peas>>>>>> Juice>>>>> Juice>>>> DNA (rating 1- 5)

#1>>>>>> Frozen>>>>> Unfiltered >3 ml>>>> 1

#2>>>>>> Frozen>>>>> Filtered>>> 1 ml>>>> 1

#3>>>>>> Dried>>>>>> Unfiltered>> 1 ml>>>> 4

#4>>>>>> Dried>>>>>> Filtered>>>> 3 ml>>>> 5

Thanks to a "free gift" from DELL, I'm the owner of what must be the crappiest digital camera in the universe, so bear with me on the image below. The tubes are arranged for left to right as #1, #2, #3 and #4. The photo below was taken approximately 5 min after adding the methylated spirits. Although the images make it difficult to discern the amount of DNA separated, I used a visual (direct, non-photographical) ranking system as my response variable, on a scale from 1 to 5, with ranking 5 indicating the highest amount of DNA and 1 the lowest. This system is of course subjective and imprecise but at the time it was the best I could do: the amounts would not lend themselves to weighing with my electronic scales, which only have a 0.1 g precision.



In tube #1, an opaque disc of DNA could be seen, risen to the top of the alcohol phase. In tube#2, a similarly faint disc was about to rise to the top too. I ranked both as 1.

For tubes #3 and #4, please note that the foam mentioned before had settled on top of the alcohol phase and this foam is not indicative of the amount of DNA. But between the pea soup/alcohol separation and the foam, it can be clearly seen that much more DNA has gathered there than in the case of runs #1 and #2. I ranked #3 as a 4 and #4 as a 5.

To calculate the effects of the factors, simply add the response variables (DNA) of both runs where the factor was at one level (1), add the response variables (DNA) of both runs where the factor was at the other level (2), calculate (2) - (1) and divide by 2.

For example in the case of the type of peas, (1) = 1 + 1 = 2, (2) = 4 + 5 = 9, (2) - (1) = 7 and the effect of switching from frozen to dried peas thus equals 7/2 = +3.5.

Similarly the effect of switching from unfiltered to filtered pineapple juice was +0.5 and the effect of switching from 1 ml to 3 ml of pineapple juice was -0.5. The average of all four runs was 2.75.

So, let's draw some conclusions:

1. Peas: the effect of the type of peas, predictably perhaps, was very clear; dried peas yield more of the stringy stuff, presumably because they contain more cell matter and less water.


2. Filtering: due to the low resolution of the design, an effect value of +0.5 is probably not statistically significant. It's also in breach of a previous assumption, namely that using pineapple cell pulp, instead of filtered pineapple juice, would introduce more DNA to the mix and increase yield. It's entirely possible though that using unfiltered pineapple juice in conjunction with a DNA-poor cell soup (i.e. using frozen peas) could slightly increase the yield of DNA (as was previously observed).


3. Amount of pineapple juice: again, due to the low resolution of the design, an effect value of -0.5 is probably not statistically significant. In practice this would mean that the required amount of pineapple juice is probably less than 1 ml and that a few drops would do the trick. Enzymes are bio-catalysts and very effective at what they do: a small amount probably suffices.

And so, I'll probably be running one more test, this time using split peas and determining the minimum needed quantity of filtered pineapple juice...