Email about Plate Miniprep vs Standard Miniprep

2015-01-13

azim58 - Email about Plate Miniprep vs Standard Miniprep

The final decision after this e-mail was basically to do the 100 minipreps

===========================================================================

We can meet if you want too, but I'll try to get most of the pertinent
information in this e-mail. Basically, we have a tough decision to make
about how to split up our scFv PCANTAB culture after transformation in
preparation for a miniprep. I did some calculations, and we would need an
unreasonable number of plates.

For example, let's say the transformation efficiency with the ligation is
1*10^8 cfu/ug. Note that the normal transformation efficiency of the
commercially competent DH10B cells is >1*10^10 cfu/ug so I have already
tried to compensate for the lower transformation efficiency of ligation.
About the most DNA we could have used for this transformation would be
0.7 ug.

So that would leave us with

7*10^7 cfu

If we split this up into 5,000 cfu samples there would be 14,000 samples,
which would require about 140 of those 96 well plates. This would be
extremely expensive and impractical.

One issue is that the plates are not really cheaper than a regular
miniprep.

Qiagen Regular Miniprep: $354/(250 samples) = $1.42/sample

Qiagen 96 DirectPrep: $501/(4*96) = $1.30/sample

Qiagen 96 Turboprep: $898/(4*96) = $2.34/sample
(Mark and Andrey used the Turboprep kit before)

Another issue is that Andrey says the plates do not work as well as the
regular miniprep. He says that the vacuum is weaker than centrifuging,
that the plates are more cumbersome to work with, and that larger
quantities and purer DNA are obtained from the regular miniprep. He also
said that he and Mark have even miniprepped several hundred samples at
once using the regular miniprep method.

We also need to keep in mind that I will need to repeat whatever process
I choose for all 4 groups (tumor, naive, SMC1fs, and 1-78).

So. . .

Do we even need to be so extremely concerned with library bias? What if I
just arbitrarily split the transformation culture up into 100 cultures
regardless of the transformation efficiency? If I did this, then I could
do 100 standard minipreps. If the transformation efficiency were 1*10^8
cfu/ug like I said before than the number of colonies in each of my
samples would be 7*10^5. If we choose the plate method, with 140 plates
we would have to spend 140/4*501 = $17,535 for just one of my four groups
which would be very impractical.

So what do you think? With my inexperience, I'm leaning towards just
simply doing 100 regular minipreps. It works well, it's cheap, people are
already used to doing minipreps, etc. We would just be trading off a
little bit on the complexity.

azim58wiki: